WO2021212654A1 - Physical machine resource allocation model acquisition method and apparatus, and computer device - Google Patents

Physical machine resource allocation model acquisition method and apparatus, and computer device Download PDF

Info

Publication number
WO2021212654A1
WO2021212654A1 PCT/CN2020/098788 CN2020098788W WO2021212654A1 WO 2021212654 A1 WO2021212654 A1 WO 2021212654A1 CN 2020098788 W CN2020098788 W CN 2020098788W WO 2021212654 A1 WO2021212654 A1 WO 2021212654A1
Authority
WO
WIPO (PCT)
Prior art keywords
physical machine
historical
resource allocation
hidden layer
training
Prior art date
Application number
PCT/CN2020/098788
Other languages
French (fr)
Chinese (zh)
Inventor
田玉凯
Original Assignee
平安科技（深圳）有限公司
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 平安科技（深圳）有限公司 filed Critical 平安科技（深圳）有限公司
Publication of WO2021212654A1 publication Critical patent/WO2021212654A1/en

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06F—ELECTRIC DIGITAL DATA PROCESSING
    • G06F9/00—Arrangements for program control, e.g. control units
    • G06F9/06—Arrangements for program control, e.g. control units using stored programs, i.e. using an internal store of processing equipment to receive or retain programs
    • G06F9/46—Multiprogramming arrangements
    • G06F9/50—Allocation of resources, e.g. of the central processing unit [CPU]
    • G06F9/5005—Allocation of resources, e.g. of the central processing unit [CPU] to service a request
    • G06F9/5027—Allocation of resources, e.g. of the central processing unit [CPU] to service a request the resource being a machine, e.g. CPUs, Servers, Terminals
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/045—Combinations of networks
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/04—Architecture, e.g. interconnection topology
    • G06N3/049—Temporal neural networks, e.g. delay elements, oscillating neurons or pulsed inputs
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks
    • G06N3/08—Learning methods
    • G06N3/084—Backpropagation, e.g. using gradient descent

Definitions

  • This application relates to the field of artificial intelligence, in particular to a method, device and computer equipment for acquiring a physical machine resource allocation model.
  • the main purpose of this application is to provide a method, device and computer equipment for acquiring a physical machine resource allocation model, which aims to solve the technical problem that the prior art cannot accurately predict the physical machine resource allocation.
  • this application proposes a method for acquiring a physical machine resource allocation model, which includes:
  • Preprocess the historical usage data to obtain usable historical data, and divide the historical data into a training data set and a test data set;
  • the training data in the training data set is input into the pre-hidden layer and post-hidden layer in the preset BILSTM neural network in the forward and reverse order of time, respectively, for training, and the output value of the previous term is obtained.
  • the training data is summed and then averaged to obtain the average training value, and the average training value is input into the antecedent hidden layer and the latter hidden layer in the BILSTM neural network for training, and the corresponding The output value X of the hidden layer;
  • a basic model of physical machine resource allocation based on the BILSTM neural network is written, and the basic model is trained to obtain the physical machine resource allocation model, wherein the physical machine resource allocation
  • This application also provides a device for acquiring a physical machine resource allocation model, including:
  • the obtaining unit is used to obtain the historical usage data of the physical machine in the specified area
  • a processing unit configured to preprocess the historical usage data to obtain usable historical data, and divide the historical data into a training data set and a test data set;
  • the second training unit is used for summing the training data and then averaging processing to obtain an average training value, and inputting the average training value into the preceding hidden layer and the succeeding hidden layer in the BILSTM neural network Training in, get the output value X of the corresponding hidden layer;
  • the first calculation unit is configured to calculate the product of the hidden layer output Hi of each historical node and the probability weight according to the probability distribution of the output value X on each historical node, and then add the product to obtain the value C;
  • the training unit for writing a basic model of physical machine resource allocation based on the BILSTM neural network according to the hidden layer output value hk', and training the basic model to obtain the physical machine resource allocation model, wherein:
  • the present application also provides a computer device, including a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, a method for acquiring a resource allocation model of a physical machine is implemented.
  • the method includes the following steps :
  • Preprocess the historical usage data to obtain usable historical data, and divide the historical data into a training data set and a test data set;
  • the training data in the training data set is input into the pre-hidden layer and post-hidden layer in the preset BILSTM neural network in the forward and reverse order of time, respectively, for training, and the output value of the previous term is obtained.
  • the training data is summed and then averaged to obtain the average training value, and the average training value is input into the antecedent hidden layer and the latter hidden layer in the BILSTM neural network for training, and the corresponding The output value X of the hidden layer;
  • a basic model of physical machine resource allocation based on the BILSTM neural network is written, and the basic model is trained to obtain the physical machine resource allocation model, wherein the physical machine resource allocation
  • the present application also provides a computer-readable storage medium on which a computer program is stored.
  • a method for acquiring a resource allocation model of a physical machine is implemented. The method includes the following steps:
  • Preprocess the historical usage data to obtain usable historical data, and divide the historical data into a training data set and a test data set;
  • the training data in the training data set is input into the pre-hidden layer and post-hidden layer in the preset BILSTM neural network in the forward and reverse order of time, respectively, for training, and the output value of the previous term is obtained.
  • the training data is summed and then averaged to obtain the average training value, and the average training value is input into the antecedent hidden layer and the latter hidden layer in the BILSTM neural network for training, and the corresponding The output value X of the hidden layer;
  • a basic model of physical machine resource allocation based on the BILSTM neural network is written, and the basic model is trained to obtain the physical machine resource allocation model, wherein the physical machine resource allocation
  • This application uses the attention mechanism and
  • FIG. 1 is a schematic flowchart of a method for acquiring a physical machine resource allocation model according to an embodiment of the application
  • FIG. 2 is a schematic block diagram of the structure of an apparatus for acquiring a resource allocation model of a physical machine according to an embodiment of the application;
  • FIG. 3 is a schematic block diagram of the structure of a computer device according to an embodiment of the application.
  • an embodiment of the present application provides a method for acquiring a physical machine resource allocation model, including:
  • the training data is summed and then averaged to obtain an average training value, and the average training value is input into the antecedent hidden layer and the latter hidden layer in the BILSTM neural network for training to obtain The output value X of the hidden layer;
  • the executors that execute the above methods are generally electronic computer equipment with computing capabilities such as servers and computer equipment.
  • the above physical machine refers to a physical device such as a server.
  • the aforementioned historical usage data generally includes data such as GUP (Graphic Processing Unit, Chinese translated as "graphic processor") data, hard disk capacity data, network card rate data, and physical machine usage data.
  • GUP Graphic Processing Unit, Chinese translated as "graphic processor”
  • the above-mentioned designated area refers to a geographical area or a network environment range, for example, physical machine usage data in a certain area, or physical machine usage data in a network environment corresponding to a certain business field.
  • the above preprocessing is the process of data cleaning.
  • the data that does not meet the following preset BILSTM neural network for training is deleted or deformed, such as normalizing irregular data and text
  • the attribute data undergoes one-hot transformation, etc., which will not be repeated here.
  • Randomly distribute the cleaned historical data to obtain a training data set and a test data set.
  • the data in the training data set is used to train the following BILSTM neural network
  • the data in the test data set is used to test the trained BILSTM neural network it's usable or not.
  • the above-mentioned BILSTM (Bi-directional Long Short-Term Memory is a combination of forward LSTM and backward LSTM) neural network is a two-way input LSTM neural network, that is, a
  • the node has an antecedent hidden layer and a posterior hidden layer.
  • the data needs to be input into the antecedent hidden layer and the posterior hidden layer in chronological order, forward and backward, and then the antecedent The output of the hidden layer and the subsequent hidden layer are added to obtain the hidden layer output of the node.
  • This application selects the BILSTM neural network for training, which can improve the data impact between various time points, thereby improving the subsequent prediction accuracy.
  • the training data is trained, but also the training data is added and averaged to obtain an average training value, and a node corresponding to the average training value is set for training, and the hidden layer output value corresponding to the node is obtained .
  • the training data is data generated within 12 months, and each month's data corresponds to a node.
  • the BILSTM neural network has 13 summary points, and the extra node is the node corresponding to the average training value. Adding the average training value to the training process can improve the calculation of the subsequent attention mechanism and improve the prediction accuracy.
  • the above step S5 of calculating the product of the hidden layer output Hi of each historical node and the probability weight based on the probability distribution of the output value X on each historical node, and then adding the value C to obtain the step S5 includes :
  • Ski vtanh(Whk+Uhi+b), k is a historical node, i is a historical node, and b is a constant;
  • Ski vtanh(Whk+Uhi+b) is one of the basic calculation formulas in the LSTM neural network, and will not be explained here.
  • the above-mentioned probability distribution of the output value X on each historical node refers to the influence of the previous node on X, that is, the weight distribution.
  • the function of the value C is to subsequently calculate the hidden layer output value hk' corresponding to the output value X.
  • the above step S2 of preprocessing the historical usage data to obtain usable historical data, and dividing the historical data into a training data set and a test data set includes:
  • Classify the historical usage data where the classification includes the zoom type that needs to be reduced, the normalization type that needs to be normalized, and the text type that needs one-hot conversion;
  • the available historical data is divided into the training data set and the test data set.
  • the data of the above-mentioned zoom type is only data whose attribute data is greater than the preset value, which is reduced in proportion to reduce the impact of the training result due to the large gap between the data;
  • the above-mentioned normalized type refers to irregular Data, such as discrete data in historical usage data, etc., normalize it to between 0 and 1, to prevent overfitting during model training later;
  • the above text types refer to data that cannot be directly trained, and one will be used. -hot transformation, so that neural network training can be carried out.
  • the method of dividing the available historical data into the training data set and the test data set may be to divide the available historical data in each time period according to 7/3, or 8/ The ratio of 2 is randomly divided to obtain the training data and test data in each time period. For example, 12 months of historical data, each month is a time period, then 70% of the time period in the first month is randomly selected As the test data, the remaining 30% of the data is used as the test data, and so on.
  • the training data in each time period basically maintains a sufficient amount of training to prevent all or most of the data in a certain time period from being allocated to the test data set when the overall random allocation is performed, thereby affecting the training of the neural network Results etc.
  • the step S1 of obtaining historical usage data of a physical machine in a designated area includes:
  • the historical usage data is not obtained blindly, but selectively obtained. For example, first determine the number of nodes of the BILSTM neural network, and know how many historical usage data needs to be included according to the number of nodes The time span, specifically, because there is one more node used to calculate the average training value in this application, the number of nodes needs to be reduced by one. For example, the number of nodes in the BILSTM neural network is 13, which should correspond to different 12 consecutive For the data of the time span, the last node is used to train the average of the data in the 12 time span. After determining the specific number of nodes, the start time and end time of historical data can be determined according to the above-mentioned time span.
  • the start time is the time corresponding to the current time before December, and the current time is the end time.
  • the start time will be the month time corresponding to the current time removed, and the end time of the month before the current time is the historical data
  • the end time, and the start time of the historical data is 12 months forward from the end time to get the time.
  • the above step of compiling a basic model of physical machine resource allocation based on the BILSTM neural network according to the hidden layer output value hk', and training the basic model to obtain the physical machine resource allocation model S7 including:
  • the main purpose is to train multiple physical machine resource allocation models with different training levels to facilitate later use.
  • three physical machine resource allocation models can be trained.
  • a physical machine resource allocation model trained using training data in all training data sets has the highest prediction accuracy. It is defined as an advanced physical machine resource allocation model.
  • the model obtained with the medium amount of data is defined as the middle-level physical machine resource allocation model, and the model obtained with the least amount of training data is defined as the low-level physical machine resource allocation model.
  • the model can be used according to the corresponding level of the model user's identity.
  • the model user is the company's decision maker, and it uses the advanced physical machine resource allocation model, which can accurately grasp the physical machine resource allocation Others, such as customer service personnel, who need a simple understanding of the prediction results to respond to customer questions, use low-level physical machine resource allocation models, etc., to prevent accurate data from leaking to competitors.
  • the advanced physical machine resource allocation model which can accurately grasp the physical machine resource allocation Others, such as customer service personnel, who need a simple understanding of the prediction results to respond to customer questions, use low-level physical machine resource allocation models, etc., to prevent accurate data from leaking to competitors.
  • the above step of compiling a basic model of physical machine resource allocation based on the BILSTM neural network according to the hidden layer output value hk', and training the basic model to obtain the physical machine resource allocation model After that, include:
  • start the identity recognition program to obtain the identity of the current operator
  • the physical machine resource allocation model registered corresponding to the identity of the operator is allocated.
  • the above-mentioned identity recognition program can be any program in the prior art, such as fingerprint recognition, face recognition, voiceprint recognition, and so on. Different operators can call different levels of physical machine resource allocation models, which can effectively control the security of resource allocation data.
  • This application uses the attention
  • this application also provides a device for acquiring a physical machine resource allocation model, including:
  • the acquiring unit 10 is configured to acquire historical usage data of physical machines in a designated area;
  • the processing unit 20 is configured to preprocess the historical usage data to obtain usable historical data, and divide the historical data into a training data set and a test data set;
  • the second training unit 40 is used for summing the training data and then averaging processing to obtain an average training value, and inputting the average training value into the antecedent hidden layer and the latter hidden layer in the BILSTM neural network Perform training in the layer to obtain the output value X of the corresponding hidden layer;
  • the first calculation unit 50 is configured to calculate the product of the hidden layer output Hi of each historical node and the probability weight according to the probability distribution of the output value X on each historical node, and then add the product to obtain the value C;
  • the writing training unit 70 is used for writing a basic model of physical machine resource allocation based on the BILSTM neural network according to the hidden layer output value hk', and training the basic model to obtain the physical machine resource allocation model, wherein ,
  • the foregoing first calculation unit 50 includes:
  • the construction module is used to construct a formula according to the probability distribution of the output value X on each historical node:
  • Ski vtanh(Whk+Uhi+b), k is a historical node, i is a historical node, and b is a constant;
  • the aforementioned processing unit 20 includes:
  • the classification module is used to classify the historical usage data, where the classification includes the zoom type that needs to be reduced, the normalized type that needs to be normalized, and the text type that needs one-hot conversion;
  • a processing module configured to process the classified data according to a corresponding processing method to obtain the available historical data
  • the splitting module is used to divide the available historical data into the training data set and the test data set.
  • the aforementioned processing unit 20 includes:
  • the first obtaining module is configured to obtain the preset number of nodes of the BILSTM neural network
  • a determining module configured to determine the start time and end time of the historical data according to the value of the number of nodes minus one;
  • the second acquisition module is configured to acquire the data between the start time and the end time as the historical usage data.
  • the above-mentioned writing training unit 70 includes:
  • a writing module for writing a basic model of physical machine resource allocation based on the BILSTM neural network according to the hidden layer output value hk';
  • the training module is used to train the basic model using different amounts of training data in the training data set to obtain the physical machine resource allocation models of different training levels, wherein the less training data is used, the obtained The lower the prediction accuracy of the physical machine resource allocation model.
  • the acquisition device of the above-mentioned physical machine resource allocation model includes:
  • An acquiring command unit for acquiring a command for calling the resource allocation model of the physical machine
  • the identity recognition unit is used to start the identity recognition program according to the command to obtain the identity of the current operator
  • the allocation model unit is used to allocate the physical machine resource allocation model registered corresponding to the operator's identity according to the operator's identity.
  • the above-mentioned units and modules are devices that execute the above-mentioned method steps, and will not be described here.
  • an embodiment of the present application also provides a computer device.
  • the computer device may be a server, and its internal structure may be as shown in FIG. 3.
  • the computer equipment includes a processor, a memory, a network interface, and a database connected through a system bus. Among them, the processor designed by the computer is used to provide calculation and control capabilities.
  • the memory of the computer device includes a non-volatile storage medium and an internal memory.
  • the non-volatile storage medium stores an operating system, a computer program, and a database.
  • the memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium.
  • the database of the computer equipment is used for data such as historical usage data.
  • the network interface of the computer device is used to communicate with an external terminal through a network connection.
  • the computer program is executed by the processor to implement the method for acquiring the physical machine resource allocation model in any of the above embodiments.
  • FIG. 3 is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied.
  • the embodiments of the present application also provide a computer-readable storage medium.
  • the computer-readable storage medium may be non-volatile or volatile, and has a computer program stored thereon.
  • the computer program is executed by a processor, the foregoing A method for acquiring a physical machine resource allocation model in any embodiment.
  • Non-volatile memory may include read only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory.
  • Volatile memory may include random access memory (RAM) or external cache memory.
  • RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual-rate data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous Link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

Landscapes

  • Engineering & Computer Science (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • Software Systems (AREA)
  • General Physics & Mathematics (AREA)
  • General Engineering & Computer Science (AREA)
  • Data Mining & Analysis (AREA)
  • Computational Linguistics (AREA)
  • Biophysics (AREA)
  • Evolutionary Computation (AREA)
  • General Health & Medical Sciences (AREA)
  • Molecular Biology (AREA)
  • Computing Systems (AREA)
  • Biomedical Technology (AREA)
  • Artificial Intelligence (AREA)
  • Mathematical Physics (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • Health & Medical Sciences (AREA)
  • Management, Administration, Business Operations System, And Electronic Commerce (AREA)

Abstract

The present application relates to the field of artificial intelligence. Disclosed are a physical machine resource allocation model acquisition method and apparatus, and a computer device. The method comprises: inputting data in a training data set into a preset BILSTM neural network for training, and obtaining an output value Hi=(hi+hi') of a hidden layer; summing training data and then averaging same to obtain an average training value, and inputting the average training value into the BILSTM neural network for training to obtain an output value X of the corresponding hidden layer; according to probability distribution of the output value X on historical nodes, respectively calculating the product of the hidden layer output Hi of each historical node and a probability weight, and performing adding to obtain a value C; calculating, by means of a linear function Hk'=H(C, hk, Xk), a hidden layer output value hk' corresponding to the output value X; and writing, according to the hidden layer output value hk', a physical machine resource allocation basic model based on the BILSTM neural network, and training the basic model to obtain a physical machine resource allocation model. The present application improves prediction precision.

Description

物理机资源分配模型的获取方法、装置和计算机设备Method, device and computer equipment for acquiring physical machine resource allocation model
本申请要求于2020年4月24日提交中国专利局、申请号为202010334968.3，发明名称为“物理机资源分配模型的获取方法、装置和计算机设备”的中国专利申请的优先权，其全部内容通过引用结合在本申请中。This application claims the priority of a Chinese patent application filed with the Chinese Patent Office on April 24, 2020, the application number is 202010334968.3, and the invention title is "Methods, Apparatus, and Computer Equipment for Obtaining Physical Machine Resource Allocation Models". The entire content of this application is approved. The reference is incorporated in this application.
技术领域Technical field
本申请涉及到人工智能领域，特别是涉及到一种物理机资源分配模型的获取方法、装置和计算机设备。This application relates to the field of artificial intelligence, in particular to a method, device and computer equipment for acquiring a physical machine resource allocation model.
背景技术Background technique
传统的物理机资源分配主要是是根据经验分配，发明人发现，往往会出现各个区的物理机资源准备不能满足业务的需求，有时物理机资源准备充分却没有市场，有时物理机资源没有准备充分确有很大的市场需求，从而导致资源利用不充分。Traditional physical machine resource allocation is mainly based on experience. The inventor found that it often happens that the physical machine resource preparation in each area cannot meet the needs of the business. Sometimes the physical machine resources are fully prepared but there is no market, and sometimes the physical machine resources are not fully prepared. There is indeed a huge market demand, which leads to insufficient utilization of resources.
所以亟需一种可以准确预测物理机资源分配的方法。Therefore, a method that can accurately predict the resource allocation of a physical machine is urgently needed.
技术问题technical problem
本申请的主要目的为提供一种物理机资源分配模型的获取方法、装置和计算机设备，旨在解决现有技术无法准确预测物理机资源分配的技术问题。The main purpose of this application is to provide a method, device and computer equipment for acquiring a physical machine resource allocation model, which aims to solve the technical problem that the prior art cannot accurately predict the physical machine resource allocation.
技术解决方案Technical solutions
为了实现上述发明目的，本申请提出一种物理机资源分配模型的获取方法，包括：In order to achieve the above-mentioned purpose of the invention, this application proposes a method for acquiring a physical machine resource allocation model, which includes:
获取指定区域的物理机的历史使用数据；Obtain historical usage data of physical machines in a specified area;
将所述历史使用数据进行预处理后得到可用的历史数据，并将所述历史数据分成训练数据集和测试数据集；Preprocess the historical usage data to obtain usable historical data, and divide the historical data into a training data set and a test data set;
将所述训练数据集中的训练数据按照时间的正向顺序和反向顺序分别输入到预设的BILSTM神经网络中的前项隐含层和后项隐含层中进行训练，得到前项输出值hi和后项输出值hi’，并且将所述hi与hi’相加得到隐含层的输出值Hi＝(hi+hi’)，其中i为某个历史节点；以及，The training data in the training data set is input into the pre-hidden layer and post-hidden layer in the preset BILSTM neural network in the forward and reverse order of time, respectively, for training, and the output value of the previous term is obtained. hi and the subsequent term output value hi', and add the hi and hi' to obtain the output value Hi=(hi+hi') of the hidden layer, where i is a certain historical node; and,
将所述训练数据进行求和后再平均处理，得到平均训练值，并将所述平均训练值输入到BILSTM神经网络中的前项隐含层和后项隐含层中进行训练，得到对应的隐含层的输出值X；The training data is summed and then averaged to obtain the average training value, and the average training value is input into the antecedent hidden layer and the latter hidden layer in the BILSTM neural network for training, and the corresponding The output value X of the hidden layer;
根据所述输出值X在各历史节点上的概率分布，计算出各历史节点的隐含层输出Hi和概率权重的乘积，然后相加得到值C；According to the probability distribution of the output value X on each historical node, calculate the product of the hidden layer output Hi of each historical node and the probability weight, and then add the product to obtain the value C;
通过线性函数Hk’＝H(C,hk，Xk)求出所述输出值X对应的隐含层输出值hk’；Calculate the hidden layer output value hk' corresponding to the output value X by linear function Hk'=H(C,hk,Xk);
依据所述隐含层输出值hk’编写基于BILSTM神经网络的物理机资源分配的基础模型，并对所述基础模型进行训练，得到所述物理机资源分配模型，其中，所述物理机资源分配模型计算出的预测值为Y＝R(X,hk’)，R为常量。According to the hidden layer output value hk', a basic model of physical machine resource allocation based on the BILSTM neural network is written, and the basic model is trained to obtain the physical machine resource allocation model, wherein the physical machine resource allocation The predicted value calculated by the model is Y=R(X,hk'), and R is a constant.
本申请还提供一种物理机资源分配模型的获取装置，包括：This application also provides a device for acquiring a physical machine resource allocation model, including:
获取单元，用于获取指定区域的物理机的历史使用数据；The obtaining unit is used to obtain the historical usage data of the physical machine in the specified area;
处理单元，用于将所述历史使用数据进行预处理后得到可用的历史数据，并将所述历史数据分成训练数据集和测试数据集；A processing unit, configured to preprocess the historical usage data to obtain usable historical data, and divide the historical data into a training data set and a test data set;
第一训练单元，用于将所述训练数据集中的训练数据按照时间的正向顺序和反向顺序分别输入到预设的BILSTM神经网络中的前项隐含层和后项隐含层中进行训练，得到前项输出值hi和后项输出值hi’，并且将所述hi与hi’相加得到隐含层的输出值Hi＝(hi+hi’)，其中i为某个历史节点；以及，The first training unit is used to input the training data in the training data set into the preceding hidden layer and the succeeding hidden layer in the preset BILSTM neural network according to the forward and reverse order of time. Training to obtain the output value hi of the previous term and the output value hi' of the latter term, and add the hi and hi' to obtain the output value Hi=(hi+hi') of the hidden layer, where i is a certain historical node; as well as,
第二训练单元，用于将所述训练数据进行求和后再平均处理，得到平均训练值，并将所述平均训练值输入到BILSTM神经网络中的前项隐含层和后项隐含层中进行训练，得到对应的隐含层的输出值X；The second training unit is used for summing the training data and then averaging processing to obtain an average training value, and inputting the average training value into the preceding hidden layer and the succeeding hidden layer in the BILSTM neural network Training in, get the output value X of the corresponding hidden layer;
第一计算单元，用于根据所述输出值X在各历史节点上的概率分布，计算出各历史节点的隐含层输出Hi和概率权重的乘积，然后相加得到值C；The first calculation unit is configured to calculate the product of the hidden layer output Hi of each historical node and the probability weight according to the probability distribution of the output value X on each historical node, and then add the product to obtain the value C;
第二计算单元，用于通过线性函数Hk’＝H(C,hk，Xk)求出所述输出值X对应的隐含层输出值hk’；The second calculation unit is used to obtain the hidden layer output value hk' corresponding to the output value X through a linear function Hk'=H(C,hk,Xk);
编写训练单元，用于依据所述隐含层输出值hk’编写基于BILSTM神经网络的物理机资源分配的基础模型，并对所述基础模型进行训练，得到所述物理机资源分配模型，其中，所述物理机资源分配模型计算出的预测值为Y＝R(X,hk’)，R为常量。Writing a training unit for writing a basic model of physical machine resource allocation based on the BILSTM neural network according to the hidden layer output value hk', and training the basic model to obtain the physical machine resource allocation model, wherein: The predicted value calculated by the physical machine resource allocation model is Y=R(X,hk'), and R is a constant.
本申请还提供一种计算机设备，包括存储器和处理器，所述存储器存储有计算机程序，所述处理器执行所述计算机程序时实现一种物理机资源分配模型的获取方法，该方法包括以下步骤：The present application also provides a computer device, including a memory and a processor, the memory stores a computer program, and when the processor executes the computer program, a method for acquiring a resource allocation model of a physical machine is implemented. The method includes the following steps :
获取指定区域的物理机的历史使用数据；Obtain historical usage data of physical machines in a specified area;
将所述历史使用数据进行预处理后得到可用的历史数据，并将所述历史数据分成训练数据集和测试数据集；Preprocess the historical usage data to obtain usable historical data, and divide the historical data into a training data set and a test data set;
将所述训练数据集中的训练数据按照时间的正向顺序和反向顺序分别输入到预设的BILSTM神经网络中的前项隐含层和后项隐含层中进行训练，得到前项输出值hi和后项输出值hi’，并且将所述hi与hi’相加得到隐含层的输出值Hi＝(hi+hi’)，其中i为某个历史节点；以及，The training data in the training data set is input into the pre-hidden layer and post-hidden layer in the preset BILSTM neural network in the forward and reverse order of time, respectively, for training, and the output value of the previous term is obtained. hi and the subsequent term output value hi', and add the hi and hi' to obtain the output value Hi=(hi+hi') of the hidden layer, where i is a certain historical node; and,
将所述训练数据进行求和后再平均处理，得到平均训练值，并将所述平均训练值输入到BILSTM神经网络中的前项隐含层和后项隐含层中进行训练，得到对应的隐含层的输出值X；The training data is summed and then averaged to obtain the average training value, and the average training value is input into the antecedent hidden layer and the latter hidden layer in the BILSTM neural network for training, and the corresponding The output value X of the hidden layer;
根据所述输出值X在各历史节点上的概率分布，计算出各历史节点的隐含层输出Hi和概率权重的乘积，然后相加得到值C；According to the probability distribution of the output value X on each historical node, calculate the product of the hidden layer output Hi of each historical node and the probability weight, and then add the product to obtain the value C;
通过线性函数Hk’＝H(C,hk，Xk)求出所述输出值X对应的隐含层输出值hk’；Calculate the hidden layer output value hk' corresponding to the output value X by linear function Hk'=H(C,hk,Xk);
依据所述隐含层输出值hk’编写基于BILSTM神经网络的物理机资源分配的基础模型，并对所述基础模型进行训练，得到所述物理机资源分配模型，其中，所述物理机资源分配模型计算出的预测值为Y＝R(X,hk’)，R为常量。According to the hidden layer output value hk', a basic model of physical machine resource allocation based on the BILSTM neural network is written, and the basic model is trained to obtain the physical machine resource allocation model, wherein the physical machine resource allocation The predicted value calculated by the model is Y=R(X,hk'), and R is a constant.
本申请还提供一种计算机可读存储介质，其上存储有计算机程序，所述计算机程序被处理器执行时实现一种物理机资源分配模型的获取方法，该方法包括以下步骤：The present application also provides a computer-readable storage medium on which a computer program is stored. When the computer program is executed by a processor, a method for acquiring a resource allocation model of a physical machine is implemented. The method includes the following steps:
获取指定区域的物理机的历史使用数据；Obtain historical usage data of physical machines in a specified area;
将所述历史使用数据进行预处理后得到可用的历史数据，并将所述历史数据分成训练数据集和测试数据集；Preprocess the historical usage data to obtain usable historical data, and divide the historical data into a training data set and a test data set;
将所述训练数据集中的训练数据按照时间的正向顺序和反向顺序分别输入到预设的BILSTM神经网络中的前项隐含层和后项隐含层中进行训练，得到前项输出值hi和后项输出值hi’，并且将所述hi与hi’相加得到隐含层的输出值Hi＝(hi+hi’)，其中i为某个历史节点；以及，The training data in the training data set is input into the pre-hidden layer and post-hidden layer in the preset BILSTM neural network in the forward and reverse order of time, respectively, for training, and the output value of the previous term is obtained. hi and the subsequent term output value hi', and add the hi and hi' to obtain the output value Hi=(hi+hi') of the hidden layer, where i is a certain historical node; and,
将所述训练数据进行求和后再平均处理，得到平均训练值，并将所述平均训练值输入到BILSTM神经网络中的前项隐含层和后项隐含层中进行训练，得到对应的隐含层的输出值X；The training data is summed and then averaged to obtain the average training value, and the average training value is input into the antecedent hidden layer and the latter hidden layer in the BILSTM neural network for training, and the corresponding The output value X of the hidden layer;
根据所述输出值X在各历史节点上的概率分布，计算出各历史节点的隐含层输出Hi和概率权重的乘积，然后相加得到值C；According to the probability distribution of the output value X on each historical node, calculate the product of the hidden layer output Hi of each historical node and the probability weight, and then add the product to obtain the value C;
通过线性函数Hk’＝H(C,hk，Xk)求出所述输出值X对应的隐含层输出值hk’；Calculate the hidden layer output value hk' corresponding to the output value X by linear function Hk'=H(C,hk,Xk);
依据所述隐含层输出值hk’编写基于BILSTM神经网络的物理机资源分配的基础模型，并对所述基础模型进行训练，得到所述物理机资源分配模型，其中，所述物理机资源分配模型计算出的预测值为Y＝R(X,hk’)，R为常量。According to the hidden layer output value hk', a basic model of physical machine resource allocation based on the BILSTM neural network is written, and the basic model is trained to obtain the physical machine resource allocation model, wherein the physical machine resource allocation The predicted value calculated by the model is Y=R(X,hk'), and R is a constant.
有益效果Beneficial effect
本申请的物理机资源分配模型的获取方法、装置和计算机设备，将训练数据集中的数据输入到预设的BILSTM神经网络中进行训练，得到隐含层的输出值Hi＝(hi+hi’)；以及，将训练数据进行求和后再平均处理，得到平均训练值，并将平均训练值输入到BILSTM神经网络中进行训练，得到对应的隐含层的输出值X；根据输出值X在各历史节点上的概率分布，计算出各历史节点的隐含层输出Hi和概率权重的乘积，然后相加得到值C；通过线性函数Hk’＝H(C,hk，Xk)求出输出值X对应的隐含层输出值hk’；依据隐含层输出值hk’编写基于BILSTM神经网络的物理机资源分配的基础模型，并对基础模型进行训练，得到物理机资源分配模型。本申请使用了注意力机制以及BILSTM神经网络，在整个技术方案中加入注意力机制后，对算法进行了改进，从而提高了预测精度，优化了架构，提高特征值的挖掘深度。The method, device and computer equipment for acquiring the physical machine resource allocation model of the present application input the data in the training data set into the preset BILSTM neural network for training, and obtain the output value of the hidden layer Hi=(hi+hi') ; And, after the training data is summed and then averaged, the average training value is obtained, and the average training value is input into the BILSTM neural network for training to obtain the output value X of the corresponding hidden layer; according to the output value X in each The probability distribution on the historical node, calculate the product of the hidden layer output Hi of each historical node and the probability weight, and then add the value C; calculate the output value X through the linear function Hk'=H(C,hk,Xk) The corresponding hidden layer output value hk'; according to the hidden layer output value hk', a basic model of physical machine resource allocation based on the BILSTM neural network is written, and the basic model is trained to obtain a physical machine resource allocation model. This application uses the attention mechanism and the BILSTM neural network. After the attention mechanism is added to the entire technical solution, the algorithm is improved, thereby improving the prediction accuracy, optimizing the architecture, and increasing the depth of feature value mining.
附图说明Description of the drawings
图1为本申请一实施例的物理机资源分配模型的获取方法的流程示意图；FIG. 1 is a schematic flowchart of a method for acquiring a physical machine resource allocation model according to an embodiment of the application;
图2为本申请一实施例的物理机资源分配模型的获取装置的结构示意框图；2 is a schematic block diagram of the structure of an apparatus for acquiring a resource allocation model of a physical machine according to an embodiment of the application;
图3为本申请一实施例的计算机设备的结构示意框图。FIG. 3 is a schematic block diagram of the structure of a computer device according to an embodiment of the application.
本申请目的的实现、功能特点及优点将结合实施例，参照附图做进一步说明。The realization, functional characteristics, and advantages of the purpose of this application will be further described in conjunction with the embodiments and with reference to the accompanying drawings.
本发明的最佳实施方式The best mode of the present invention
为了使本申请的目的、技术方案及优点更加清楚明白，以下结合附图及实施例，对本申请进行进一步详细说明。应当理解，此处描述的具体实施例仅仅用以解释本申请，并不用于限定本申请。In order to make the purpose, technical solutions, and advantages of this application clearer and clearer, the following further describes the application in detail with reference to the accompanying drawings and embodiments. It should be understood that the specific embodiments described herein are only used to explain the present application, and are not used to limit the present application.
参照图1，本申请实施例提供一种物理机资源分配模型的获取方法，包括：1, an embodiment of the present application provides a method for acquiring a physical machine resource allocation model, including:
S1、获取指定区域的物理机的历史使用数据；S1. Obtain historical usage data of physical machines in a designated area;
S2、将所述历史使用数据进行预处理后得到可用的历史数据，并将所述历史数据分成训练数据集和测试数据集；S2. Preprocess the historical usage data to obtain usable historical data, and divide the historical data into a training data set and a test data set;
S3、将所述训练数据集中的训练数据按照时间的正向顺序和反向顺序分别输入到预设的BILSTM神经网络中的前项隐含层和后项隐含层中进行训练，得到前项输出值hi和后项输出值hi’，并且将所述hi与hi’相加得到隐含层的输出值Hi＝(hi+hi’)，其中i为某个历史节点；以及，S3. Input the training data in the training data set in the forward and reverse order of time into the preceding hidden layer and the succeeding hidden layer in the preset BILSTM neural network for training, to obtain the preceding item Output the value hi and the subsequent output value hi', and add the hi and hi' to obtain the output value Hi=(hi+hi') of the hidden layer, where i is a certain historical node; and,
S4，将所述训练数据进行求和后再平均处理，得到平均训练值，并将所述平均训练值输入到BILSTM神经网络中的前项隐含层和后项隐含层中进行训练，得到所述隐含层的输出值X；S4. The training data is summed and then averaged to obtain an average training value, and the average training value is input into the antecedent hidden layer and the latter hidden layer in the BILSTM neural network for training to obtain The output value X of the hidden layer;
S5、根据所述输出值X在各历史节点上的概率分布，计算出各历史节点的隐含层输出Hi和概率权重的乘积，然后相加得到值C；S5. Calculate the product of the hidden layer output Hi of each historical node and the probability weight according to the probability distribution of the output value X on each historical node, and then add the product to obtain a value C;
S6、通过线性函数Hk’＝H(C,hk，Xk)求出所述输出值X对应的隐含层输出值hk’；S6. Calculate the hidden layer output value hk' corresponding to the output value X through the linear function Hk'=H(C,hk,Xk);
S7、依据所述隐含层输出值hk’编写基于BILSTM神经网络的物理机资源分配的基础模型，并对所述基础模型进行训练，得到所述物理机资源分配模型，其中，所述物理机资源分配模型计算出的预测值为Y＝R(X,hk’)，R为常量。S7. Write a basic model of physical machine resource allocation based on the BILSTM neural network according to the hidden layer output value hk', and train the basic model to obtain the physical machine resource allocation model, wherein the physical machine The predicted value calculated by the resource allocation model is Y=R(X,hk'), and R is a constant.
执行上述方法的执行主体一般是服务器、计算机设备等具有运算能力的电子计算机设备。The executors that execute the above methods are generally electronic computer equipment with computing capabilities such as servers and computer equipment.
如上述步骤S1所述，上述物理机是指服务器等物理设备。上述历史使用数据一般包括GUP(Graphic Processing Unit，中文翻译为“图形处理器”)的数据、硬盘容量数据、网卡速率数据、物理机的使用数量数据等数据。上述指定区域是指一个地理范围或者网络环境范围，比如，某个地域的物理机使用数据，又或者某一业务领域对应的网络环境中物理机的使用数据等。As described in step S1 above, the above physical machine refers to a physical device such as a server. The aforementioned historical usage data generally includes data such as GUP (Graphic Processing Unit, Chinese translated as "graphic processor") data, hard disk capacity data, network card rate data, and physical machine usage data. The above-mentioned designated area refers to a geographical area or a network environment range, for example, physical machine usage data in a certain area, or physical machine usage data in a network environment corresponding to a certain business field.
如上述步骤S2所述，上述预处理即为数据清理的过程，将不符合训练下述预设的BILSTM神经网络的数据进行删除或者变形，比如对不规则的数据进行归一化处理，对文字 属性的数据进行one-hot转变等，在此不在一一赘述。将清洗后的历史数据随机分分配，得到训练数据集和测试数据集，其中训练数据集中的数据用于训练下述的BILSTM神经网络，而测试数据集中的数据用于测试训练后的BILSTM神经网络是否可用。As described in step S2 above, the above preprocessing is the process of data cleaning. The data that does not meet the following preset BILSTM neural network for training is deleted or deformed, such as normalizing irregular data and text The attribute data undergoes one-hot transformation, etc., which will not be repeated here. Randomly distribute the cleaned historical data to obtain a training data set and a test data set. The data in the training data set is used to train the following BILSTM neural network, and the data in the test data set is used to test the trained BILSTM neural network it's usable or not.
如上述步骤S3和S4所述，上述BILSTM(Bi-directional Long Short-Term Memory的缩写，是由前向LSTM与后向LSTM组合而成)神经网络是一种双向输入的LSTM神经网络，即一个节点具有前项隐含层和后项隐含层，在训练或预测的时候需要数据按照时间顺序，正向和反向分别输入到前项隐含层和后项隐含层，然后将前项隐含层和后项隐含层的输出相加得到该节点的隐含层输出，本申请选择BILSTM神经网络进行训练，可以提高各时间点之间的数据影响，进而提高后续的预测准确性。在本申请中，不但对训练数据进行训练，还将训练数据相加然后求平均得到一个平均训练值，并且为该平均训练值对应设置一个节点进行训练，得到该节点对应的隐含层输出值。举个例子，训练数据为12个月内产生的数据，每一个月的数据对应设置一个节点，则BILSTM神经网络的总结点数为13个，多出的一个节点即为平均训练值对应的节点。将平均训练值加入训练过程，可以提高后续的注意力机制的计算，提高预测精度。As described in the above steps S3 and S4, the above-mentioned BILSTM (Bi-directional Long Short-Term Memory is a combination of forward LSTM and backward LSTM) neural network is a two-way input LSTM neural network, that is, a The node has an antecedent hidden layer and a posterior hidden layer. During training or prediction, the data needs to be input into the antecedent hidden layer and the posterior hidden layer in chronological order, forward and backward, and then the antecedent The output of the hidden layer and the subsequent hidden layer are added to obtain the hidden layer output of the node. This application selects the BILSTM neural network for training, which can improve the data impact between various time points, thereby improving the subsequent prediction accuracy. In this application, not only the training data is trained, but also the training data is added and averaged to obtain an average training value, and a node corresponding to the average training value is set for training, and the hidden layer output value corresponding to the node is obtained . For example, the training data is data generated within 12 months, and each month's data corresponds to a node. The BILSTM neural network has 13 summary points, and the extra node is the node corresponding to the average training value. Adding the average training value to the training process can improve the calculation of the subsequent attention mechanism and improve the prediction accuracy.
如上述步骤S5至S7所述，即为加入了注意力机制后，进行一系列的计算处理后，得到一个重新编写的基础模型，然后再对该基础模型进行训练，进而得到物理机资源分配模型。需要注意的是，对上述基础模型进行训练和测试的数据仍然是上述的训练数据集和测试数据集中的数据。物理机资源分配模型计算出的预测值为Y＝R(X,hk’)，其结合了平均训练值对应的X，以及通过注意力机制计算出的hk’，大大地提高了预测精度。在本实施例中，基于BILSTM神经网络的物理机资源分配的基础模型，其损失函数按MSE(均方误差)的标准进行计算即可。As mentioned in the above steps S5 to S7, after adding the attention mechanism, a series of calculations are performed to obtain a rewritten basic model, and then the basic model is trained to obtain the physical machine resource allocation model . It should be noted that the data for training and testing the above-mentioned basic model is still the data in the above-mentioned training data set and test data set. The predicted value calculated by the physical machine resource allocation model is Y=R(X,hk'), which combines X corresponding to the average training value and hk' calculated by the attention mechanism, which greatly improves the prediction accuracy. In this embodiment, for the basic model of physical machine resource allocation based on the BILSTM neural network, its loss function can be calculated according to the MSE (Mean Square Error) standard.
在一个实施例中，上述根据所述输出值X在各历史节点上的概率分布，计算出各历史节点的隐含层输出Hi和概率权重的乘积，然后相加得到值C的步骤S5，包括：In one embodiment, the above step S5 of calculating the product of the hidden layer output Hi of each historical node and the probability weight based on the probability distribution of the output value X on each historical node, and then adding the value C to obtain the step S5 includes :
根据所述输出值X在各历史节点上的概率分布，构建公式：According to the probability distribution of the output value X on each historical node, construct a formula:
Figure PCTCN2020098788-appb-000001
其中，Ski＝vtanh(Whk+Uhi+b),k为某个历史节点，i为某个历史节点，b为常量；
Figure PCTCN2020098788-appb-000001
Among them, Ski=vtanh(Whk+Uhi+b), k is a historical node, i is a historical node, and b is a constant;
依据公式
Figure PCTCN2020098788-appb-000002
得到所述值C。 According to the formula
Figure PCTCN2020098788-appb-000002
The value C is obtained.
在本实施例中，即为具体的计算上述值C的过程，Ski＝vtanh(Whk+Uhi+b)是LSTM神经网络中的基础计算公式之一，在此不做解释。上述输出值X在各历史节点上的概率分布是指之前节点对X的影响，即权重的分配。然后构建出极大自然数公式αki，然后计算出值C。值C的作用是用于后续计算输出值X对应的隐含层输出值hk’。In this embodiment, it is a specific process of calculating the above-mentioned value C. Ski=vtanh(Whk+Uhi+b) is one of the basic calculation formulas in the LSTM neural network, and will not be explained here. The above-mentioned probability distribution of the output value X on each historical node refers to the influence of the previous node on X, that is, the weight distribution. Then construct the maximum natural number formula αki, and then calculate the value C. The function of the value C is to subsequently calculate the hidden layer output value hk' corresponding to the output value X.
在一个实施例中，上述将所述历史使用数据进行预处理后得到可用的历史数据，并将所述历史数据分成训练数据集和测试数据集的步骤S2，包括：In an embodiment, the above step S2 of preprocessing the historical usage data to obtain usable historical data, and dividing the historical data into a training data set and a test data set includes:
对所述历史使用数据进行分类，其中，分类包括需要将数据缩小的缩放类型、需要归一化处理的归一类型以及需要one-hot转变的文字类型；Classify the historical usage data, where the classification includes the zoom type that needs to be reduced, the normalization type that needs to be normalized, and the text type that needs one-hot conversion;
将分类后的数据按照对应的处理方式进行处理得到所述可用的历史数据；Processing the classified data according to the corresponding processing method to obtain the available historical data;
将所述可用的历史数据分成所述训练数据集和所述测试数据集。The available historical data is divided into the training data set and the test data set.
在本实施例中，上述缩放类型的数据只是属性数据大于预设值的数据，将其等比例缩小，从而降低因为数据之间差距太大而影响训练的结果；上述归一类型是指不规则数据，比如历史使用数据中的离散数据等，将其归一化到0到1之间，防止之后模型训练时产生过拟合的情况；上述文字类型是指不能直接训练的数据，将进行one-hot转变，从而可以进行神经网络的训练。在本实施例中，将所述可用的历史数据分成所述训练数据集和所述测试数据集的方法可以为，将每一个时间段内的可用的历史数据，按照7/3，或8/2的比例进行随机划分得到每一个时间段内的训练数据和测试数据，比如，12个月的历史数据，每 一个月为一个时间段，那么在第一个月的时间段内随机抽取7成的数据作为测试数据，剩余的3成数据作为测试数据，并以此类推。这样处理，每一个时间段内的训练数据基本保持足够多的训练量，防止整体随机分配时，将某个时间段的数据全部或大部分被分配到测试数据集中，从而影响对神经网络的训练结果等。In this embodiment, the data of the above-mentioned zoom type is only data whose attribute data is greater than the preset value, which is reduced in proportion to reduce the impact of the training result due to the large gap between the data; the above-mentioned normalized type refers to irregular Data, such as discrete data in historical usage data, etc., normalize it to between 0 and 1, to prevent overfitting during model training later; the above text types refer to data that cannot be directly trained, and one will be used. -hot transformation, so that neural network training can be carried out. In this embodiment, the method of dividing the available historical data into the training data set and the test data set may be to divide the available historical data in each time period according to 7/3, or 8/ The ratio of 2 is randomly divided to obtain the training data and test data in each time period. For example, 12 months of historical data, each month is a time period, then 70% of the time period in the first month is randomly selected As the test data, the remaining 30% of the data is used as the test data, and so on. In this way, the training data in each time period basically maintains a sufficient amount of training to prevent all or most of the data in a certain time period from being allocated to the test data set when the overall random allocation is performed, thereby affecting the training of the neural network Results etc.
在一个实施例中，上述获取指定区域的物理机的历史使用数据的步骤S1，包括：In an embodiment, the step S1 of obtaining historical usage data of a physical machine in a designated area includes:
获取所述预设的BILSTM神经网络的节点数量；Obtaining the preset number of nodes of the BILSTM neural network;
根据所述节点数量减一的值，确定所述历史数据的起始时间和结束时间；Determine the start time and end time of the historical data according to the value of the number of nodes minus one;
获取所述起始时间和结束时间之间的数据作为所述历史使用数据。Obtain the data between the start time and the end time as the historical usage data.
在本实施例中，在获取历史使用数据的时候并不会盲目的获取，而是有选择的获取，比如，先确定BILSTM神经网络的节点数量，可以根据节点数量知道历史使用数据需要包含多少个时间跨度，具体地，因为本申请中多出一个用于计算平均训练值的节点，所以需要对节点数量进行减一处理，如BILSTM神经网络的节点数量为13，其应该对应不同的12个连续的时间跨度的数据，最后一个节点用于训练这个12个时间跨度中的数据平均值。当确定具体的节点数量后，再根据上述的时间跨度既可以确定历史数据的起始时间和结束时间。比如，时间跨度是1个月，节点数量为13，那么起始时间是12月前对应当前时间的时间，当前时间为结束时间。在另一个实施例中，如果时间跨度为1个月，且为从1号开始的自然月，则开始时间将当前时间对应的月份时间去除，当前时间的前一个月的结束时间为历史数据的结束时间，而历史数据的开始时间则从结束时间向前推12个月得到时间。通过本实施例得到的历史使用数据，在进行预处理之后对预设的BILSTM神经网络进行训练的时候，数据对应各节点更加准确，从而提高训练后的物理机资源分配模型具有更加精准的预测能力。In this embodiment, the historical usage data is not obtained blindly, but selectively obtained. For example, first determine the number of nodes of the BILSTM neural network, and know how many historical usage data needs to be included according to the number of nodes The time span, specifically, because there is one more node used to calculate the average training value in this application, the number of nodes needs to be reduced by one. For example, the number of nodes in the BILSTM neural network is 13, which should correspond to different 12 consecutive For the data of the time span, the last node is used to train the average of the data in the 12 time span. After determining the specific number of nodes, the start time and end time of historical data can be determined according to the above-mentioned time span. For example, if the time span is 1 month and the number of nodes is 13, then the start time is the time corresponding to the current time before December, and the current time is the end time. In another embodiment, if the time span is 1 month, and it is a natural month beginning on the 1st, then the start time will be the month time corresponding to the current time removed, and the end time of the month before the current time is the historical data The end time, and the start time of the historical data is 12 months forward from the end time to get the time. Through the historical usage data obtained in this embodiment, when the preset BILSTM neural network is trained after preprocessing, the data corresponds to each node more accurately, thereby improving the physical machine resource allocation model after training to have more accurate prediction capabilities .
在一个实施例中，上述依据所述隐含层输出值hk’编写基于BILSTM神经网络的物理机资源分配的基础模型，并对所述基础模型进行训练，得到所述物理机资源分配模型的步骤S7，包括：In one embodiment, the above step of compiling a basic model of physical machine resource allocation based on the BILSTM neural network according to the hidden layer output value hk', and training the basic model to obtain the physical machine resource allocation model S7, including:
依据所述隐含层输出值hk’编写基于BILSTM神经网络的物理机资源分配的基础模型；Compile a basic model of physical machine resource allocation based on BILSTM neural network according to the hidden layer output value hk';
使用所述训练数据集中不同数量的训练数据对所述基础模型进行训练，得到不同训练等级的所述物理机资源分配模型，其中，使用的训练数据越少，得到的所述物理机资源分配模型的预测准确度越低。Use different amounts of training data in the training data set to train the basic model to obtain the physical machine resource allocation models of different training levels, wherein the less training data is used, the physical machine resource allocation model is obtained The lower the accuracy of the forecast.
在本实施例中，主要是要训练得到多个训练程度不同的物理机资源分配模型，以便于后期的使用。本申请中，可以训练出三个物理机资源分配模型，如使用全部训练数据集中的训练数据训练得到的物理机资源分配模型，其预测准确性最高，定义为高级物理机资源分配模型，使用训练数据的数量中等得到的模型定义为中级述物理机资源分配模型，使用训练数据的数量最少得到的模型定义为低级述物理机资源分配模型。在使用过程，可以根据模型使用者的身份，给以对应级别的模型使用，比如，模型使用者是公司的决策者，其使用高级物理机资源分配模型，其可以准确的掌握物理机的资源分配量；而其他如客服人员等需要简单了解预测结果以回复客户问题的，则使用低级物理机资源分配模型等，以防止将准确的数据透漏给竞争对手等。In this embodiment, the main purpose is to train multiple physical machine resource allocation models with different training levels to facilitate later use. In this application, three physical machine resource allocation models can be trained. For example, a physical machine resource allocation model trained using training data in all training data sets has the highest prediction accuracy. It is defined as an advanced physical machine resource allocation model. The model obtained with the medium amount of data is defined as the middle-level physical machine resource allocation model, and the model obtained with the least amount of training data is defined as the low-level physical machine resource allocation model. In the use process, the model can be used according to the corresponding level of the model user's identity. For example, the model user is the company's decision maker, and it uses the advanced physical machine resource allocation model, which can accurately grasp the physical machine resource allocation Others, such as customer service personnel, who need a simple understanding of the prediction results to respond to customer questions, use low-level physical machine resource allocation models, etc., to prevent accurate data from leaking to competitors.
在一个实施例中，上述依据所述隐含层输出值hk’编写基于BILSTM神经网络的物理机资源分配的基础模型，并对所述基础模型进行训练，得到所述物理机资源分配模型的步骤之后，包括：In one embodiment, the above step of compiling a basic model of physical machine resource allocation based on the BILSTM neural network according to the hidden layer output value hk', and training the basic model to obtain the physical machine resource allocation model After that, include:
获取调用所述物理机资源分配模型的命令；Acquiring a command for invoking the physical machine resource allocation model;
依据所述命令，启动身份识别程序，得到当前操作者的身份；According to the command, start the identity recognition program to obtain the identity of the current operator;
根据所述操作者的身份分配与所述操作者的身份相对应登记的所述物理机资源分配模型。According to the identity of the operator, the physical machine resource allocation model registered corresponding to the identity of the operator is allocated.
在本实施例中，上述身份识别程序可以现有技术中的任意一种程序，比如指纹识别、人脸识别、声纹识别等等。不同的操作者可以调用的物理机资源分配模型的等级不同，可 以有效管控资源分配数据的安全性。In this embodiment, the above-mentioned identity recognition program can be any program in the prior art, such as fingerprint recognition, face recognition, voiceprint recognition, and so on. Different operators can call different levels of physical machine resource allocation models, which can effectively control the security of resource allocation data.
本申请的物理机资源分配模型的获取方法，将训练数据集中的数据输入到预设的BILSTM神经网络中进行训练，得到隐含层的输出值Hi＝(hi+hi’)；以及，将训练数据进行求和后再平均处理，得到平均训练值，并将平均训练值输入到BILSTM神经网络中进行训练，得到对应的隐含层的输出值X；根据输出值X在各历史节点上的概率分布，计算出各历史节点的隐含层输出Hi和概率权重的乘积，然后相加得到值C；通过线性函数Hk’＝H(C,hk，Xk)求出输出值X对应的隐含层输出值hk’；依据隐含层输出值hk’编写基于BILSTM神经网络的物理机资源分配的基础模型，并对基础模型进行训练，得到物理机资源分配模型。本申请使用了注意力机制以及BILSTM神经网络，在整个技术方案中加入注意力机制后，对算法进行了改进，从而提高了预测精度，优化了架构，提高特征值的挖掘深度。In the method for acquiring the physical machine resource allocation model of the present application, the data in the training data set is input into a preset BILSTM neural network for training, and the output value Hi=(hi+hi') of the hidden layer is obtained; and the training is performed After the data is summed and then averaged, the average training value is obtained, and the average training value is input into the BILSTM neural network for training, and the output value X of the corresponding hidden layer is obtained; according to the probability of the output value X on each historical node Distribution, calculate the product of the hidden layer output Hi of each historical node and the probability weight, and then add the value C; use the linear function Hk'=H(C,hk,Xk) to find the hidden layer corresponding to the output value X Output value hk'; according to the hidden layer output value hk', a basic model of physical machine resource allocation based on the BILSTM neural network is written, and the basic model is trained to obtain a physical machine resource allocation model. This application uses the attention mechanism and the BILSTM neural network. After the attention mechanism is added to the entire technical solution, the algorithm is improved, thereby improving the prediction accuracy, optimizing the architecture, and increasing the depth of feature value mining.
参照图2，本申请还提供一种物理机资源分配模型的获取装置，包括：2, this application also provides a device for acquiring a physical machine resource allocation model, including:
获取单元10，用于获取指定区域的物理机的历史使用数据；The acquiring unit 10 is configured to acquire historical usage data of physical machines in a designated area;
处理单元20，用于将所述历史使用数据进行预处理后得到可用的历史数据，并将所述历史数据分成训练数据集和测试数据集；The processing unit 20 is configured to preprocess the historical usage data to obtain usable historical data, and divide the historical data into a training data set and a test data set;
第一训练单元30，用于将所述训练数据集中的训练数据按照时间的正向顺序和反向顺序分别输入到预设的BILSTM神经网络中的前项隐含层和后项隐含层中进行训练，得到前项输出值hi和后项输出值hi’，并且将所述hi与hi’相加得到隐含层的输出值Hi＝(hi+hi’)，其中i为某个历史节点；以及，The first training unit 30 is configured to input the training data in the training data set into the preceding hidden layer and the succeeding hidden layer in the preset BILSTM neural network according to the forward order and the reverse order of time. Perform training to obtain the output value of the preceding term hi and the output value of the following term hi', and add the hi and hi' to obtain the output value of the hidden layer Hi=(hi+hi'), where i is a certain historical node ;as well as,
第二训练单元40，用于将所述训练数据进行求和后再平均处理，得到平均训练值，并将所述平均训练值输入到BILSTM神经网络中的前项隐含层和后项隐含层中进行训练，得到对应的隐含层的输出值X；The second training unit 40 is used for summing the training data and then averaging processing to obtain an average training value, and inputting the average training value into the antecedent hidden layer and the latter hidden layer in the BILSTM neural network Perform training in the layer to obtain the output value X of the corresponding hidden layer;
第一计算单元50，用于根据所述输出值X在各历史节点上的概率分布，计算出各历史节点的隐含层输出Hi和概率权重的乘积，然后相加得到值C；The first calculation unit 50 is configured to calculate the product of the hidden layer output Hi of each historical node and the probability weight according to the probability distribution of the output value X on each historical node, and then add the product to obtain the value C;
第二计算单元60，用于通过线性函数Hk’＝H(C,hk，Xk)求出所述输出值X对应的隐含层输出值hk’；The second calculation unit 60 is configured to obtain the hidden layer output value hk' corresponding to the output value X through a linear function Hk'=H(C,hk,Xk);
编写训练单元70，用于依据所述隐含层输出值hk’编写基于BILSTM神经网络的物理机资源分配的基础模型，并对所述基础模型进行训练，得到所述物理机资源分配模型，其中，所述物理机资源分配模型计算出的预测值为Y＝R(X,hk’)，R为常量。The writing training unit 70 is used for writing a basic model of physical machine resource allocation based on the BILSTM neural network according to the hidden layer output value hk', and training the basic model to obtain the physical machine resource allocation model, wherein , The predicted value calculated by the physical machine resource allocation model is Y=R(X,hk'), and R is a constant.
在一个实施例中，上述第一计算单元50，包括：In an embodiment, the foregoing first calculation unit 50 includes:
构建模块，用于根据所述输出值X在各历史节点上的概率分布，构建公式：The construction module is used to construct a formula according to the probability distribution of the output value X on each historical node:
Figure PCTCN2020098788-appb-000003
其中，Ski＝vtanh(Whk+Uhi+b),k为某个历史节点，i为某个历史节点，b为常量；
Figure PCTCN2020098788-appb-000003
Among them, Ski=vtanh(Whk+Uhi+b), k is a historical node, i is a historical node, and b is a constant;
计算模块，用于依据公式
Figure PCTCN2020098788-appb-000004
得到所述值C。 Calculation module for formula
Figure PCTCN2020098788-appb-000004
The value C is obtained.
在一个实施例中，上述处理单元20，包括：In an embodiment, the aforementioned processing unit 20 includes:
分类模块，用于对所述历史使用数据进行分类，其中，分类包括需要将数据缩小的缩放类型、需要归一化处理的归一类型以及需要one-hot转变的文字类型；The classification module is used to classify the historical usage data, where the classification includes the zoom type that needs to be reduced, the normalized type that needs to be normalized, and the text type that needs one-hot conversion;
处理模块，用于将分类后的数据按照对应的处理方式进行处理得到所述可用的历史数据；A processing module, configured to process the classified data according to a corresponding processing method to obtain the available historical data;
拆分模块，用于将所述可用的历史数据分成所述训练数据集和所述测试数据集。The splitting module is used to divide the available historical data into the training data set and the test data set.
在一个实施例中，上述处理单元20，包括：In an embodiment, the aforementioned processing unit 20 includes:
第一获取模块，用于获取所述预设的BILSTM神经网络的节点数量；The first obtaining module is configured to obtain the preset number of nodes of the BILSTM neural network;
确定模块，用于根据所述节点数量减一的值，确定所述历史数据的起始时间和结束时间；A determining module, configured to determine the start time and end time of the historical data according to the value of the number of nodes minus one;
第二获取模块，用于获取所述起始时间和结束时间之间的数据作为所述历史使用数据。The second acquisition module is configured to acquire the data between the start time and the end time as the historical usage data.
在一个实施例中，上述编写训练单元70，包括：In an embodiment, the above-mentioned writing training unit 70 includes:
编写模块，用于依据所述隐含层输出值hk’编写基于BILSTM神经网络的物理机资源分配的基础模型；A writing module for writing a basic model of physical machine resource allocation based on the BILSTM neural network according to the hidden layer output value hk';
训练模块，用于使用所述训练数据集中不同数量的训练数据对所述基础模型进行训练，得到不同训练等级的所述物理机资源分配模型，其中，使用的训练数据越少，得到的所述物理机资源分配模型的预测准确度越低。The training module is used to train the basic model using different amounts of training data in the training data set to obtain the physical machine resource allocation models of different training levels, wherein the less training data is used, the obtained The lower the prediction accuracy of the physical machine resource allocation model.
进一步地，上述物理机资源分配模型的获取装置，包括：Further, the acquisition device of the above-mentioned physical machine resource allocation model includes:
获取命令单元，用于获取调用所述物理机资源分配模型的命令；An acquiring command unit for acquiring a command for calling the resource allocation model of the physical machine;
身份识别单元，用于依据所述命令，启动身份识别程序，得到当前操作者的身份；The identity recognition unit is used to start the identity recognition program according to the command to obtain the identity of the current operator;
分配模型单元，用于根据所述操作者的身份分配与所述操作者的身份相对应登记的所述物理机资源分配模型。The allocation model unit is used to allocate the physical machine resource allocation model registered corresponding to the operator's identity according to the operator's identity.
上述各单元和模块是执行上述各方法步骤的装置，在此不在展开一一说明。The above-mentioned units and modules are devices that execute the above-mentioned method steps, and will not be described here.
参照图3，本申请实施例中还提供一种计算机设备，该计算机设备可以是服务器，其内部结构可以如图3所示。该计算机设备包括通过***总线连接的处理器、存储器、网络接口和数据库。其中，该计算机设计的处理器用于提供计算和控制能力。该计算机设备的存储器包括非易失性存储介质、内存储器。该非易失性存储介质存储有操作***、计算机程序和数据库。该内存器为非易失性存储介质中的操作***和计算机程序的运行提供环境。该计算机设备的数据库用于历史使用数据等数据。该计算机设备的网络接口用于与外部的终端通过网络连接通信。该计算机程序被处理器执行时以实现上述任一实施例中的物理机资源分配模型的获取方法。Referring to FIG. 3, an embodiment of the present application also provides a computer device. The computer device may be a server, and its internal structure may be as shown in FIG. 3. The computer equipment includes a processor, a memory, a network interface, and a database connected through a system bus. Among them, the processor designed by the computer is used to provide calculation and control capabilities. The memory of the computer device includes a non-volatile storage medium and an internal memory. The non-volatile storage medium stores an operating system, a computer program, and a database. The memory provides an environment for the operation of the operating system and the computer program in the non-volatile storage medium. The database of the computer equipment is used for data such as historical usage data. The network interface of the computer device is used to communicate with an external terminal through a network connection. The computer program is executed by the processor to implement the method for acquiring the physical machine resource allocation model in any of the above embodiments.
本领域技术人员可以理解，图3中示出的结构，仅仅是与本申请方案相关的部分结构的框图，并不构成对本申请方案所应用于其上的计算机设备的限定。Those skilled in the art can understand that the structure shown in FIG. 3 is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied.
本申请实施例还提供一种计算机可读存储介质，所述计算机可读存储介质可以是非易失性，也可以是易失性，其上存储有计算机程序，计算机程序被处理器执行时实现上述任一实施例中的物理机资源分配模型的获取方法。The embodiments of the present application also provide a computer-readable storage medium. The computer-readable storage medium may be non-volatile or volatile, and has a computer program stored thereon. When the computer program is executed by a processor, the foregoing A method for acquiring a physical machine resource allocation model in any embodiment.
本领域普通技术人员可以理解实现上述实施例方法中的全部或部分流程，是可以通过计算机程序来指令相关的硬件来完成，所述的计算机程序可存储与一非易失性计算机可读取存储介质中，该计算机程序在执行时，可包括如上述各方法的实施例的流程。其中，本申请所提供的和实施例中所使用的对存储器、存储、数据库或其它介质的任何引用，均可包括非易失性和/或易失性存储器。非易失性存储器可以包括只读存储器(ROM)、可编程ROM(PROM)、电可编程ROM(EPROM)、电可擦除可编程ROM(EEPROM)或闪存。易失性存储器可包括随机存取存储器(RAM)或者外部高速缓冲存储器。作为说明而非局限，RAM一多种形式可得，诸如静态RAM(SRAM)、动态RAM(DRAM)、同步DRAM(SDRAM)、双速据率SDRAM(SSRSDRAM)、增强型SDRAM(ESDRAM)、同步链路(Synchlink)DRAM(SLDRAM)、存储器总线(Rambus)直接RAM(RDRAM)、直接存储器总线动态RAM(DRDRAM)、以及存储器总线动态RAM(RDRAM)等。A person of ordinary skill in the art can understand that all or part of the processes in the above-mentioned embodiment methods can be implemented by instructing relevant hardware through a computer program. The computer program can be stored and a non-volatile computer readable storage. In the medium, when the computer program is executed, it may include the processes of the above-mentioned method embodiments. Wherein, any reference to memory, storage, database or other media provided in this application and used in the embodiments may include non-volatile and/or volatile memory. Non-volatile memory may include read only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. As an illustration and not a limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual-rate data rate SDRAM (SSRSDRAM), enhanced SDRAM (ESDRAM), synchronous Link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.
需要说明的是，在本文中，术语“包括”、“包含”或者其任何其他变体意在涵盖非排他性的包含，从而使得包括一系列要素的过程、装置、物品或者方法不仅包括那些要素，而且还包括没有明确列出的其他要素，或者是还包括为这种过程、装置、物品或者方法所固有的要素。在没有更多限制的情况下，由语句“包括一个……”限定的要素，并不排除在包括该要素的过程、装置、物品或者方法中还存在另外的相同要素。It should be noted that in this article, the terms "include", "include" or any other variants thereof are intended to cover non-exclusive inclusion, so that a process, device, article or method including a series of elements not only includes those elements, It also includes other elements not explicitly listed, or elements inherent to the process, device, article, or method. If there are no more restrictions, the element defined by the sentence "including a..." does not exclude the existence of other identical elements in the process, device, article, or method that includes the element.
以上所述仅为本申请的优选实施例，并非因此限制本申请的专利范围，凡是利用本申请说明书及附图内容所作的等效结构或等效流程变换，或直接或间接运用在其他相关的技术领域，均同理包括在本申请的专利保护范围内。The above are only the preferred embodiments of this application, and do not limit the scope of this application. Any equivalent structure or equivalent process transformation made by using the content of the description and drawings of this application, or directly or indirectly applied to other related The technical field is equally included in the scope of patent protection of this application.

Claims (20)

  1. 一种物理机资源分配模型的获取方法，其中，包括：A method for acquiring a physical machine resource allocation model, which includes:
    获取指定区域的物理机的历史使用数据；Obtain historical usage data of physical machines in a specified area;
    将所述历史使用数据进行预处理后得到可用的历史数据，并将所述历史数据分成训练数据集和测试数据集；Preprocess the historical usage data to obtain usable historical data, and divide the historical data into a training data set and a test data set;
    将所述训练数据集中的训练数据按照时间的正向顺序和反向顺序分别输入到预设的BILSTM神经网络中的前项隐含层和后项隐含层中进行训练，得到前项输出值hi和后项输出值hi’，并且将所述hi与hi’相加得到隐含层的输出值Hi＝(hi+hi’)，其中i为某个历史节点；以及，The training data in the training data set is input into the pre-hidden layer and post-hidden layer in the preset BILSTM neural network in the forward and reverse order of time, respectively, for training, and the output value of the previous term is obtained. hi and the subsequent term output value hi', and add the hi and hi' to obtain the output value Hi=(hi+hi') of the hidden layer, where i is a certain historical node; and,
    将所述训练数据进行求和后再平均处理，得到平均训练值，并将所述平均训练值输入到BILSTM神经网络中的前项隐含层和后项隐含层中进行训练，得到对应的隐含层的输出值X；The training data is summed and then averaged to obtain the average training value, and the average training value is input into the antecedent hidden layer and the latter hidden layer in the BILSTM neural network for training, and the corresponding The output value X of the hidden layer;
    根据所述输出值X在各历史节点上的概率分布，计算出各历史节点的隐含层输出Hi和概率权重的乘积，然后相加得到值C；According to the probability distribution of the output value X on each historical node, calculate the product of the hidden layer output Hi of each historical node and the probability weight, and then add the product to obtain the value C;
    通过线性函数Hk’＝H(C,hk，Xk)求出所述输出值X对应的隐含层输出值hk’；Calculate the hidden layer output value hk' corresponding to the output value X by linear function Hk'=H(C,hk,Xk);
    依据所述隐含层输出值hk’编写基于BILSTM神经网络的物理机资源分配的基础模型，并对所述基础模型进行训练，得到所述物理机资源分配模型，其中，所述物理机资源分配模型计算出的预测值为Y＝R(X,hk’)，R为常量。According to the hidden layer output value hk', a basic model of physical machine resource allocation based on the BILSTM neural network is written, and the basic model is trained to obtain the physical machine resource allocation model, wherein the physical machine resource allocation The predicted value calculated by the model is Y=R(X,hk'), and R is a constant.
  2. 根据权利要求1所述的物理机资源分配模型的获取方法，其中，所述根据所述输出值X在各历史节点上的概率分布，计算出各历史节点的隐含层输出Hi和概率权重的乘积，然后相加得到值C的步骤，包括：The method for acquiring a resource allocation model of a physical machine according to claim 1, wherein the calculation of the hidden layer output Hi and the probability weight of each historical node according to the probability distribution of the output value X on each historical node The steps of multiplying the product and then adding up to get the value C include:
    根据所述输出值X在各历史节点上的概率分布，构建公式：According to the probability distribution of the output value X on each historical node, construct a formula:
    Figure PCTCN2020098788-appb-100001
    其中，Ski＝vtanh(Whk+Uhi+b),k为某个历史节点，i为某个历史节点，b为常量；
    Figure PCTCN2020098788-appb-100001
    Among them, Ski=vtanh(Whk+Uhi+b), k is a historical node, i is a historical node, and b is a constant;
    依据公式
    Figure PCTCN2020098788-appb-100002
    得到所述值C。     According to the formula
    Figure PCTCN2020098788-appb-100002
    The value C is obtained.
  3. 根据权利要求1所述的物理机资源分配模型的获取方法，其中，所述将所述历史使用数据进行预处理后得到可用的历史数据，并将所述历史数据分成训练数据集和测试数据集的步骤，包括：The method for acquiring a resource allocation model of a physical machine according to claim 1, wherein the historical use data is preprocessed to obtain usable historical data, and the historical data is divided into a training data set and a test data set The steps include:
    对所述历史使用数据进行分类，其中，分类包括需要将数据缩小的缩放类型、需要归一化处理的归一类型以及需要one-hot转变的文字类型；Classify the historical usage data, where the classification includes the zoom type that needs to be reduced, the normalization type that needs to be normalized, and the text type that needs one-hot conversion;
    将分类后的数据按照对应的处理方式进行处理得到所述可用的历史数据；Processing the classified data according to the corresponding processing method to obtain the available historical data;
    将所述可用的历史数据分成所述训练数据集和所述测试数据集。The available historical data is divided into the training data set and the test data set.
  4. 根据权利要求1所述的物理机资源分配模型的获取方法，其中，所述获取指定区域的物理机的历史使用数据的步骤，包括：The method for acquiring a resource allocation model of a physical machine according to claim 1, wherein the step of acquiring historical usage data of a physical machine in a designated area comprises:
    获取所述预设的BILSTM神经网络的节点数量；Obtaining the preset number of nodes of the BILSTM neural network;
    根据所述节点数量减一的值，确定所述历史数据的起始时间和结束时间；Determine the start time and end time of the historical data according to the value of the number of nodes minus one;
    获取所述起始时间和结束时间之间的数据作为所述历史使用数据。Obtain the data between the start time and the end time as the historical usage data.
  5. 根据权利要求1所述的物理机资源分配模型的获取方法，其中，所述依据所述隐含层输出值hk’编写基于BILSTM神经网络的物理机资源分配的基础模型，并对所述基础模型进行训练，得到所述物理机资源分配模型的步骤，包括：The method for acquiring a physical machine resource allocation model according to claim 1, wherein the basic model of physical machine resource allocation based on the BILSTM neural network is written according to the hidden layer output value hk', and the basic model The step of training to obtain the physical machine resource allocation model includes:
    依据所述隐含层输出值hk’编写基于BILSTM神经网络的物理机资源分配的基础模型；Compile a basic model of physical machine resource allocation based on BILSTM neural network according to the hidden layer output value hk';
    使用所述训练数据集中不同数量的训练数据对所述基础模型进行训练，得到不同训练等级的所述物理机资源分配模型，其中，使用的训练数据越少，得到的所述物理机资源分 配模型的预测准确度越低。Use different amounts of training data in the training data set to train the basic model to obtain the physical machine resource allocation models of different training levels, wherein the less training data is used, the physical machine resource allocation model is obtained The lower the accuracy of the forecast.
  6. 根据权利要求5所述的物理机资源分配模型的获取方法，其中，所述依据所述隐含层输出值hk’编写基于BILSTM神经网络的物理机资源分配的基础模型，并对所述基础模型进行训练，得到所述物理机资源分配模型的步骤之后，包括：The method for acquiring a physical machine resource allocation model according to claim 5, wherein the basic model of physical machine resource allocation based on the BILSTM neural network is written according to the hidden layer output value hk', and the basic model After training, the step of obtaining the physical machine resource allocation model includes:
    获取调用所述物理机资源分配模型的命令；Acquiring a command for invoking the physical machine resource allocation model;
    依据所述命令，启动身份识别程序，得到当前操作者的身份；According to the command, start the identity recognition program to obtain the identity of the current operator;
    根据所述操作者的身份分配与所述操作者的身份相对应登记的所述物理机资源分配模型。According to the identity of the operator, the physical machine resource allocation model registered corresponding to the identity of the operator is allocated.
  7. 一种物理机资源分配模型的获取装置，其中，包括：A device for acquiring a resource allocation model of a physical machine, which includes:
    获取单元，用于获取指定区域的物理机的历史使用数据；The obtaining unit is used to obtain the historical usage data of the physical machine in the specified area;
    处理单元，用于将所述历史使用数据进行预处理后得到可用的历史数据，并将所述历史数据分成训练数据集和测试数据集；A processing unit, configured to preprocess the historical usage data to obtain usable historical data, and divide the historical data into a training data set and a test data set;
    第一训练单元，用于将所述训练数据集中的训练数据按照时间的正向顺序和反向顺序分别输入到预设的BILSTM神经网络中的前项隐含层和后项隐含层中进行训练，得到前项输出值hi和后项输出值hi’，并且将所述hi与hi’相加得到隐含层的输出值Hi＝(hi+hi’)，其中i为某个历史节点；以及，The first training unit is used to input the training data in the training data set into the preceding hidden layer and the succeeding hidden layer in the preset BILSTM neural network according to the forward and reverse order of time. Training to obtain the output value hi of the previous term and the output value hi' of the latter term, and add the hi and hi' to obtain the output value Hi=(hi+hi') of the hidden layer, where i is a certain historical node; as well as,
    第二训练单元，用于将所述训练数据进行求和后再平均处理，得到平均训练值，并将所述平均训练值输入到LSTM神经网络中的前项隐含层和后项隐含层中进行训练，得到对应的隐含层的输出值X；The second training unit is used for summing the training data and then averaging processing to obtain an average training value, and inputting the average training value into the preceding hidden layer and the succeeding hidden layer in the LSTM neural network Training in, get the output value X of the corresponding hidden layer;
    第一计算单元，用于根据所述输出值X在各历史节点上的概率分布，计算出各历史节点的隐含层输出Hi和概率权重的乘积，然后相加得到值C；The first calculation unit is configured to calculate the product of the hidden layer output Hi of each historical node and the probability weight according to the probability distribution of the output value X on each historical node, and then add the product to obtain the value C;
    第二计算单元，用于通过线性函数Hk’＝H(C,hk，Xk)求出所述输出值X对应的隐含层输出值hk’；The second calculation unit is used to obtain the hidden layer output value hk' corresponding to the output value X through a linear function Hk'=H(C,hk,Xk);
    编写训练单元，用于依据所述隐含层输出值hk’编写基于BILSTM神经网络的物理机资源分配的基础模型，并对所述基础模型进行训练，得到所述物理机资源分配模型，其中，所述物理机资源分配模型计算出的预测值为Y＝R(X,hk’)，R为常量。Writing a training unit for writing a basic model of physical machine resource allocation based on the BILSTM neural network according to the hidden layer output value hk', and training the basic model to obtain the physical machine resource allocation model, wherein: The predicted value calculated by the physical machine resource allocation model is Y=R(X,hk'), and R is a constant.
  8. 根据权利要求7所述的物理机资源分配模型的获取装置，其中，所述第一计算单元，包括：The apparatus for acquiring a resource allocation model of a physical machine according to claim 7, wherein the first calculation unit comprises:
    构建模块，用于根据所述输出值X在各历史节点上的概率分布，构建公式：The construction module is used to construct a formula according to the probability distribution of the output value X on each historical node:
    Figure PCTCN2020098788-appb-100003
    其中，Ski＝vtanh(Whk+Uhi+b),k为某个历史节点，i为某个历史节点，b为常量；
    Figure PCTCN2020098788-appb-100003
    Among them, Ski=vtanh(Whk+Uhi+b), k is a historical node, i is a historical node, and b is a constant;
    计算模块，用于依据公式
    Figure PCTCN2020098788-appb-100004
    得到所述值C。     Calculation module for formula
    Figure PCTCN2020098788-appb-100004
    The value C is obtained.
  9. 一种计算机设备，包括存储器和处理器，所述存储器存储有计算机程序，其中，所述处理器执行所述计算机程序时实现一种物理机资源分配模型的获取方法，该方法包括以下步骤：A computer device includes a memory and a processor, and the memory stores a computer program. When the processor executes the computer program, a method for acquiring a resource allocation model of a physical machine is implemented. The method includes the following steps:
    获取指定区域的物理机的历史使用数据；Obtain historical usage data of physical machines in a specified area;
    将所述历史使用数据进行预处理后得到可用的历史数据，并将所述历史数据分成训练数据集和测试数据集；Preprocess the historical usage data to obtain usable historical data, and divide the historical data into a training data set and a test data set;
    将所述训练数据集中的训练数据按照时间的正向顺序和反向顺序分别输入到预设的BILSTM神经网络中的前项隐含层和后项隐含层中进行训练，得到前项输出值hi和后项输出值hi’，并且将所述hi与hi’相加得到隐含层的输出值Hi＝(hi+hi’)，其中i为某个历史节点；以及，The training data in the training data set is input into the pre-hidden layer and post-hidden layer in the preset BILSTM neural network in the forward and reverse order of time, respectively, for training, and the output value of the previous term is obtained. hi and the subsequent term output value hi', and add the hi and hi' to obtain the output value Hi=(hi+hi') of the hidden layer, where i is a certain historical node; and,
    将所述训练数据进行求和后再平均处理，得到平均训练值，并将所述平均训练值输入到BILSTM神经网络中的前项隐含层和后项隐含层中进行训练，得到对应的隐含层的输出值 X；The training data is summed and then averaged to obtain the average training value, and the average training value is input into the antecedent hidden layer and the latter hidden layer in the BILSTM neural network for training, and the corresponding The output value X of the hidden layer;
    根据所述输出值X在各历史节点上的概率分布，计算出各历史节点的隐含层输出Hi和概率权重的乘积，然后相加得到值C；According to the probability distribution of the output value X on each historical node, calculate the product of the hidden layer output Hi of each historical node and the probability weight, and then add the product to obtain the value C;
    通过线性函数Hk’＝H(C,hk，Xk)求出所述输出值X对应的隐含层输出值hk’；Calculate the hidden layer output value hk' corresponding to the output value X by linear function Hk'=H(C,hk,Xk);
    依据所述隐含层输出值hk’编写基于BILSTM神经网络的物理机资源分配的基础模型，并对所述基础模型进行训练，得到所述物理机资源分配模型，其中，所述物理机资源分配模型计算出的预测值为Y＝R(X,hk’)，R为常量。According to the hidden layer output value hk', a basic model of physical machine resource allocation based on the BILSTM neural network is written, and the basic model is trained to obtain the physical machine resource allocation model, wherein the physical machine resource allocation The predicted value calculated by the model is Y=R(X,hk'), and R is a constant.
  10. 根据权利要求9所述的计算机设备，其中，所述根据所述输出值X在各历史节点上的概率分布，计算出各历史节点的隐含层输出Hi和概率权重的乘积，然后相加得到值C的步骤，包括：The computer device according to claim 9, wherein the product of the hidden layer output Hi of each historical node and the probability weight is calculated according to the probability distribution of the output value X on each historical node, and then added to obtain The steps for value C include:
    根据所述输出值X在各历史节点上的概率分布，构建公式：According to the probability distribution of the output value X on each historical node, construct a formula:
    Figure PCTCN2020098788-appb-100005
    其中，Ski＝vtanh(Whk+Uhi+b),k为某个历史节点，i为某个历史节点，b为常量；
    Figure PCTCN2020098788-appb-100005
    Among them, Ski=vtanh(Whk+Uhi+b), k is a historical node, i is a historical node, and b is a constant;
    依据公式
    Figure PCTCN2020098788-appb-100006
    得到所述值C。     According to the formula
    Figure PCTCN2020098788-appb-100006
    The value C is obtained.
  11. 根据权利要求9所述的计算机设备，其中，所述将所述历史使用数据进行预处理后得到可用的历史数据，并将所述历史数据分成训练数据集和测试数据集的步骤，包括：The computer device according to claim 9, wherein the step of preprocessing the historical usage data to obtain usable historical data, and dividing the historical data into a training data set and a test data set, comprises:
    对所述历史使用数据进行分类，其中，分类包括需要将数据缩小的缩放类型、需要归一化处理的归一类型以及需要one-hot转变的文字类型；Classify the historical usage data, where the classification includes the zoom type that needs to be reduced, the normalization type that needs to be normalized, and the text type that needs one-hot conversion;
    将分类后的数据按照对应的处理方式进行处理得到所述可用的历史数据；Processing the classified data according to the corresponding processing method to obtain the available historical data;
    将所述可用的历史数据分成所述训练数据集和所述测试数据集。The available historical data is divided into the training data set and the test data set.
  12. 根据权利要求9所述的计算机设备，其中，所述获取指定区域的物理机的历史使用数据的步骤，包括：9. The computer device according to claim 9, wherein the step of obtaining historical usage data of a physical machine in a designated area comprises:
    获取所述预设的BILSTM神经网络的节点数量；Obtaining the preset number of nodes of the BILSTM neural network;
    根据所述节点数量减一的值，确定所述历史数据的起始时间和结束时间；Determine the start time and end time of the historical data according to the value of the number of nodes minus one;
    获取所述起始时间和结束时间之间的数据作为所述历史使用数据。Obtain the data between the start time and the end time as the historical usage data.
  13. 根据权利要求9所述的计算机设备，其中，所述依据所述隐含层输出值hk’编写基于BILSTM神经网络的物理机资源分配的基础模型，并对所述基础模型进行训练，得到所述物理机资源分配模型的步骤，包括：The computer device according to claim 9, wherein the basic model of physical machine resource allocation based on BILSTM neural network is written according to the hidden layer output value hk', and the basic model is trained to obtain the The steps of the physical machine resource allocation model include:
    依据所述隐含层输出值hk’编写基于BILSTM神经网络的物理机资源分配的基础模型；Compile a basic model of physical machine resource allocation based on BILSTM neural network according to the hidden layer output value hk';
    使用所述训练数据集中不同数量的训练数据对所述基础模型进行训练，得到不同训练等级的所述物理机资源分配模型，其中，使用的训练数据越少，得到的所述物理机资源分配模型的预测准确度越低。Use different amounts of training data in the training data set to train the basic model to obtain the physical machine resource allocation models of different training levels, wherein the less training data is used, the physical machine resource allocation model is obtained The lower the accuracy of the forecast.
  14. 根据权利要求13所述的计算机设备，其中，所述依据所述隐含层输出值hk’编写基于BILSTM神经网络的物理机资源分配的基础模型，并对所述基础模型进行训练，得到所述物理机资源分配模型的步骤之后，包括：The computer device according to claim 13, wherein the basic model of physical machine resource allocation based on the BILSTM neural network is written according to the hidden layer output value hk', and the basic model is trained to obtain the After the steps of the physical machine resource allocation model, include:
    获取调用所述物理机资源分配模型的命令；Acquiring a command for invoking the physical machine resource allocation model;
    依据所述命令，启动身份识别程序，得到当前操作者的身份；According to the command, start the identity recognition program to obtain the identity of the current operator;
    根据所述操作者的身份分配与所述操作者的身份相对应登记的所述物理机资源分配模型。According to the identity of the operator, the physical machine resource allocation model registered corresponding to the identity of the operator is allocated.
  15. 一种计算机可读存储介质，其上存储有计算机程序，其中，所述计算机程序被处理器执行时实现一种物理机资源分配模型的获取方法，该方法包括以下步骤：A computer-readable storage medium having a computer program stored thereon, wherein when the computer program is executed by a processor, a method for acquiring a resource allocation model of a physical machine is implemented, and the method includes the following steps:
    获取指定区域的物理机的历史使用数据；Obtain historical usage data of physical machines in a specified area;
    将所述历史使用数据进行预处理后得到可用的历史数据，并将所述历史数据分成训练数据集和测试数据集；Preprocess the historical usage data to obtain usable historical data, and divide the historical data into a training data set and a test data set;
    将所述训练数据集中的训练数据按照时间的正向顺序和反向顺序分别输入到预设的BILSTM神经网络中的前项隐含层和后项隐含层中进行训练，得到前项输出值hi和后项输出值hi’，并且将所述hi与hi’相加得到隐含层的输出值Hi＝(hi+hi’)，其中i为某个历史节点；以及，The training data in the training data set is input into the pre-hidden layer and post-hidden layer in the preset BILSTM neural network in the forward and reverse order of time, respectively, for training, and the output value of the previous term is obtained. hi and the subsequent term output value hi', and add the hi and hi' to obtain the output value Hi=(hi+hi') of the hidden layer, where i is a certain historical node; and,
    将所述训练数据进行求和后再平均处理，得到平均训练值，并将所述平均训练值输入到BILSTM神经网络中的前项隐含层和后项隐含层中进行训练，得到对应的隐含层的输出值X；The training data is summed and then averaged to obtain the average training value, and the average training value is input into the antecedent hidden layer and the latter hidden layer in the BILSTM neural network for training, and the corresponding The output value X of the hidden layer;
    根据所述输出值X在各历史节点上的概率分布，计算出各历史节点的隐含层输出Hi和概率权重的乘积，然后相加得到值C；According to the probability distribution of the output value X on each historical node, calculate the product of the hidden layer output Hi of each historical node and the probability weight, and then add the product to obtain the value C;
    通过线性函数Hk’＝H(C,hk，Xk)求出所述输出值X对应的隐含层输出值hk’；Calculate the hidden layer output value hk' corresponding to the output value X by linear function Hk'=H(C,hk,Xk);
    依据所述隐含层输出值hk’编写基于BILSTM神经网络的物理机资源分配的基础模型，并对所述基础模型进行训练，得到所述物理机资源分配模型，其中，所述物理机资源分配模型计算出的预测值为Y＝R(X,hk’)，R为常量。According to the hidden layer output value hk', a basic model of physical machine resource allocation based on the BILSTM neural network is written, and the basic model is trained to obtain the physical machine resource allocation model, wherein the physical machine resource allocation The predicted value calculated by the model is Y=R(X,hk'), and R is a constant.
  16. 根据权利要求15所述的计算机设备，其中，所述根据所述输出值X在各历史节点上的概率分布，计算出各历史节点的隐含层输出Hi和概率权重的乘积，然后相加得到值C的步骤，包括：The computer device according to claim 15, wherein the product of the hidden layer output Hi of each historical node and the probability weight is calculated according to the probability distribution of the output value X on each historical node, and then added to obtain The steps for value C include:
    根据所述输出值X在各历史节点上的概率分布，构建公式：According to the probability distribution of the output value X on each historical node, construct a formula:
    Figure PCTCN2020098788-appb-100007
    其中，Ski＝vtanh(Whk+Uhi+b),k为某个历史节点，i为某个历史节点，b为常量；
    Figure PCTCN2020098788-appb-100007
    Among them, Ski=vtanh(Whk+Uhi+b), k is a historical node, i is a historical node, and b is a constant;
    依据公式
    Figure PCTCN2020098788-appb-100008
    得到所述值C。     According to the formula
    Figure PCTCN2020098788-appb-100008
    The value C is obtained.
  17. 根据权利要求15所述的计算机可读存储介质，其中，所述将所述历史使用数据进行预处理后得到可用的历史数据，并将所述历史数据分成训练数据集和测试数据集的步骤，包括：The computer-readable storage medium according to claim 15, wherein the step of preprocessing the historical usage data to obtain usable historical data, and dividing the historical data into a training data set and a test data set, include:
    对所述历史使用数据进行分类，其中，分类包括需要将数据缩小的缩放类型、需要归一化处理的归一类型以及需要one-hot转变的文字类型；Classify the historical usage data, where the classification includes the zoom type that needs to be reduced, the normalization type that needs to be normalized, and the text type that needs one-hot conversion;
    将分类后的数据按照对应的处理方式进行处理得到所述可用的历史数据；Processing the classified data according to the corresponding processing method to obtain the available historical data;
    将所述可用的历史数据分成所述训练数据集和所述测试数据集。The available historical data is divided into the training data set and the test data set.
  18. 根据权利要求15所述的计算机可读存储介质，其中，所述获取指定区域的物理机的历史使用数据的步骤，包括：15. The computer-readable storage medium according to claim 15, wherein the step of obtaining historical usage data of a physical machine in a designated area comprises:
    获取所述预设的BILSTM神经网络的节点数量；Obtaining the preset number of nodes of the BILSTM neural network;
    根据所述节点数量减一的值，确定所述历史数据的起始时间和结束时间；Determine the start time and end time of the historical data according to the value of the number of nodes minus one;
    获取所述起始时间和结束时间之间的数据作为所述历史使用数据。Obtain the data between the start time and the end time as the historical usage data.
  19. 根据权利要求15所述的计算机可读存储介质，其中，所述依据所述隐含层输出值hk’编写基于BILSTM神经网络的物理机资源分配的基础模型，并对所述基础模型进行训练，得到所述物理机资源分配模型的步骤，包括：The computer-readable storage medium according to claim 15, wherein the basic model of physical machine resource allocation based on the BILSTM neural network is written according to the hidden layer output value hk', and the basic model is trained, The step of obtaining the physical machine resource allocation model includes:
    依据所述隐含层输出值hk’编写基于BILSTM神经网络的物理机资源分配的基础模型；Compile a basic model of physical machine resource allocation based on BILSTM neural network according to the hidden layer output value hk';
    使用所述训练数据集中不同数量的训练数据对所述基础模型进行训练，得到不同训练等级的所述物理机资源分配模型，其中，使用的训练数据越少，得到的所述物理机资源分配模型的预测准确度越低。Use different amounts of training data in the training data set to train the basic model to obtain the physical machine resource allocation models of different training levels, wherein the less training data is used, the physical machine resource allocation model is obtained The lower the accuracy of the forecast.
  20. 根据权利要求19所述的计算机可读存储介质，其中，所述依据所述隐含层输出值hk’编写基于BILSTM神经网络的物理机资源分配的基础模型，并对所述基础模型进行训练，得到所述物理机资源分配模型的步骤之后，包括：The computer-readable storage medium according to claim 19, wherein the basic model of physical machine resource allocation based on the BILSTM neural network is written according to the hidden layer output value hk', and the basic model is trained, After the step of obtaining the physical machine resource allocation model, the method includes:
    获取调用所述物理机资源分配模型的命令；Acquiring a command for invoking the physical machine resource allocation model;
    依据所述命令，启动身份识别程序，得到当前操作者的身份；According to the command, start the identity recognition program to obtain the identity of the current operator;
    根据所述操作者的身份分配与所述操作者的身份相对应登记的所述物理机资源分配模型。According to the identity of the operator, the physical machine resource allocation model registered corresponding to the identity of the operator is allocated.
PCT/CN2020/098788 2020-04-24 2020-06-29 Physical machine resource allocation model acquisition method and apparatus, and computer device WO2021212654A1 (en)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010334968.3 2020-04-24
CN202010334968.3A CN111651264A (en) 2020-04-24 2020-04-24 Method and device for acquiring physical machine resource allocation model and computer equipment

Publications (1)

Publication Number Publication Date
WO2021212654A1 true WO2021212654A1 (en) 2021-10-28

Family

ID=72352187

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2020/098788 WO2021212654A1 (en) 2020-04-24 2020-06-29 Physical machine resource allocation model acquisition method and apparatus, and computer device

Country Status (2)

Country Link
CN (1) CN111651264A (en)
WO (1) WO2021212654A1 (en)

Cited By (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117350518A (en) * 2023-12-04 2024-01-05 深圳智慧园区信息技术有限公司 Intelligent park emergency resource scheduling method and system

Families Citing this family (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN115130929B (en) * 2022-08-29 2022-11-15 中国西安卫星测控中心 Resource pool intelligent generation method based on machine learning classification

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20190130248A1 (en) * 2017-10-27 2019-05-02 Salesforce.Com, Inc. Generating dual sequence inferences using a neural network model
CN110276638A (en) * 2019-05-29 2019-09-24 南京邮电大学 A kind of Electricity price forecasting solution and system based on two-way shot and long term neural network
CN110837683A (en) * 2019-05-20 2020-02-25 全球能源互联网研究院有限公司 Training and predicting method and device for prediction model of transient stability of power system
CN111046961A (en) * 2019-12-16 2020-04-21 浙江大学 Fault classification method based on bidirectional long-and-short-term memory unit and capsule network

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20190130248A1 (en) * 2017-10-27 2019-05-02 Salesforce.Com, Inc. Generating dual sequence inferences using a neural network model
CN110837683A (en) * 2019-05-20 2020-02-25 全球能源互联网研究院有限公司 Training and predicting method and device for prediction model of transient stability of power system
CN110276638A (en) * 2019-05-29 2019-09-24 南京邮电大学 A kind of Electricity price forecasting solution and system based on two-way shot and long term neural network
CN111046961A (en) * 2019-12-16 2020-04-21 浙江大学 Fault classification method based on bidirectional long-and-short-term memory unit and capsule network

Cited By (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN117350518A (en) * 2023-12-04 2024-01-05 深圳智慧园区信息技术有限公司 Intelligent park emergency resource scheduling method and system
CN117350518B (en) * 2023-12-04 2024-03-29 深圳智慧园区信息技术有限公司 Intelligent park emergency resource scheduling method and system

Also Published As

Publication number Publication date
CN111651264A (en) 2020-09-11

Similar Documents

Publication Publication Date Title
CN110363387B (en) Portrait analysis method and device based on big data, computer equipment and storage medium
Jones New tools in non-linear modelling and prediction
US8073729B2 (en) Forecasting discovery costs based on interpolation of historic event patterns
WO2021212654A1 (en) Physical machine resource allocation model acquisition method and apparatus, and computer device
CN110888911A (en) Sample data processing method and device, computer equipment and storage medium
CN106934413B (en) Model training method, device and system and sample set optimization method and device
CN111062486B (en) Method and device for evaluating feature distribution and confidence of data
CN112035519B (en) User image drawing method, device, computer readable storage medium and terminal equipment
CN112561364A (en) Power grid operation and maintenance cost level rationality evaluation technology based on combination weight
Minku et al. Clustering dycom: An online cross-company software effort estimation study
CN111061948B (en) User tag recommendation method and device, computer equipment and storage medium
CN108764553B (en) User scale prediction method and device and computer equipment
WO2021139255A1 (en) Model based method and apparatus for predicting data change frequency, and computer device
WO2021259003A1 (en) Feature recognition method and apparatus, and computer device and storage medium
CN111652471B (en) List distribution control method and device, electronic equipment and storage medium
GB2583176A (en) Prediction device, prediction program, and prediction method for predicting human judgments
CN116502210A (en) Account authority distribution method, device and equipment and readable storage medium
CN116188174A (en) Insurance fraud detection method and system based on modularity and mutual information
Chou Modeling daily price limits
CN111598390B (en) Method, device, equipment and readable storage medium for evaluating high availability of server
CN113689114A (en) Method, device and equipment for determining credit degree
CN112613920A (en) Loss probability prediction method and device
US20220004564A1 (en) Space partitioning method for database table, device and storage medium
Fox Course LNIRT: Modeling response accuracy and response times
CN116451190B (en) Data authority setting method based on Internet medical service system

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 20932762

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 20932762

Country of ref document: EP

Kind code of ref document: A1