CN116522002A - Container recommendation method and system of navigation service system based on machine learning - Google Patents
Container recommendation method and system of navigation service system based on machine learning Download PDFInfo
- Publication number
- CN116522002A CN116522002A CN202310767471.4A CN202310767471A CN116522002A CN 116522002 A CN116522002 A CN 116522002A CN 202310767471 A CN202310767471 A CN 202310767471A CN 116522002 A CN116522002 A CN 116522002A
- Authority
- CN
- China
- Prior art keywords
- container
- feature vector
- similarity
- containers
- historical data
- Prior art date
- Legal status (The legal status 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 status listed.)
- Granted
Links
- 238000000034 method Methods 0.000 title claims abstract description 58
- 238000010801 machine learning Methods 0.000 title claims abstract description 56
- 239000013598 vector Substances 0.000 claims abstract description 269
- 238000012545 processing Methods 0.000 claims abstract description 41
- 238000012549 training Methods 0.000 claims abstract description 34
- 238000010606 normalization Methods 0.000 claims description 4
- 238000003066 decision tree Methods 0.000 description 56
- 238000004422 calculation algorithm Methods 0.000 description 36
- 238000009826 distribution Methods 0.000 description 12
- 238000012360 testing method Methods 0.000 description 10
- 238000013145 classification model Methods 0.000 description 8
- 238000013138 pruning Methods 0.000 description 8
- 230000006870 function Effects 0.000 description 5
- 230000008569 process Effects 0.000 description 5
- 230000008901 benefit Effects 0.000 description 4
- 230000007547 defect Effects 0.000 description 4
- 238000012217 deletion Methods 0.000 description 4
- 230000037430 deletion Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 238000005516 engineering process Methods 0.000 description 4
- 238000009828 non-uniform distribution Methods 0.000 description 4
- 238000004891 communication Methods 0.000 description 2
- 230000008878 coupling Effects 0.000 description 2
- 238000010168 coupling process Methods 0.000 description 2
- 238000005859 coupling reaction Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 230000010354 integration Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000004806 packaging method and process Methods 0.000 description 2
- 238000013515 script Methods 0.000 description 2
- 230000006978 adaptation Effects 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000015556 catabolic process Effects 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000007246 mechanism Effects 0.000 description 1
- 238000013508 migration Methods 0.000 description 1
- 230000005012 migration Effects 0.000 description 1
- 238000010295 mobile communication Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000002093 peripheral effect Effects 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9535—Search customisation based on user profiles and personalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/95—Retrieval from the web
- G06F16/953—Querying, e.g. by the use of web search engines
- G06F16/9536—Search customisation based on social or collaborative filtering
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/10—Pre-processing; Data cleansing
- G06F18/15—Statistical pre-processing, e.g. techniques for normalisation or restoring missing data
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/23—Clustering techniques
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/24—Classification techniques
- G06F18/243—Classification techniques relating to the number of classes
- G06F18/24323—Tree-organised classifiers
-
- Y—GENERAL TAGGING OF NEW TECHNOLOGICAL DEVELOPMENTS; GENERAL TAGGING OF CROSS-SECTIONAL TECHNOLOGIES SPANNING OVER SEVERAL SECTIONS OF THE IPC; TECHNICAL SUBJECTS COVERED BY FORMER USPC CROSS-REFERENCE ART COLLECTIONS [XRACs] AND DIGESTS
- Y02—TECHNOLOGIES OR APPLICATIONS FOR MITIGATION OR ADAPTATION AGAINST CLIMATE CHANGE
- Y02D—CLIMATE CHANGE MITIGATION TECHNOLOGIES IN INFORMATION AND COMMUNICATION TECHNOLOGIES [ICT], I.E. INFORMATION AND COMMUNICATION TECHNOLOGIES AIMING AT THE REDUCTION OF THEIR OWN ENERGY USE
- Y02D10/00—Energy efficient computing, e.g. low power processors, power management or thermal management
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- Databases & Information Systems (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Probability & Statistics with Applications (AREA)
- Information Retrieval, Db Structures And Fs Structures Therefor (AREA)
Abstract
The invention discloses a container recommendation method and a system of a navigation service system based on machine learning, wherein the method comprises the following steps: acquiring historical data of a user using a navigation service system, and performing vectorization processing to generate a historical data feature vector; extracting a feature vector of each container, the feature vector comprising: an operational state feature vector for each container and a software functional feature vector provided by each container; modeling and training the feature vector of each container through a machine learning model, classifying the containers to obtain container types of users, container types and historical data feature vectors, setting an accurate container recommendation model, calculating first similarity between each container and the users, and recommending the container with the highest first similarity to the users; and setting a similarity recommendation model among the containers, calculating the second similarity between the container with the highest first similarity and the rest containers in the category of the container with the highest first similarity, and recommending the container with the highest second similarity to the user.
Description
Technical Field
The invention belongs to the technical field of navigation service system container recommendation, and particularly relates to a container recommendation method and system of a navigation service system based on machine learning.
Background
The greatest advantage of container technology is: 1. the greatest advantage of the quick container creation technique is that creating container instances is much faster than creating virtual machine instances, and container lightweight scripts can reduce cost in terms of performance and size. 2. The application in the fast-speed container is started directly as a process of the underlying system, not as a process inside the virtual machine. Therefore, booting the container is much faster than booting an operating system, rather than booting a process of the native device. 3. The resources occupy fewer containers, only occupy needed resources, and do not occupy unused resources; the virtual machine is a complete operating system, so that it is unavoidable to occupy all resources. In addition, multiple containers may share resources, and virtual machines are all exclusive resources. 4. The packaging small container only contains used components, and the virtual machine is the packaging of the whole operating system, so the container file is much smaller than the virtual machine file. 5. The simplified deployment adopts a container mode to deploy, the whole system can be easily combined, different services are packaged in corresponding containers through container technology, and then the containers are mutually cooperated according to requirements by combining some scripts, so that the operation can not only simplify the deployment difficulty, but also reduce the operation risk. 6. Promoting productivity containers improves the productivity of developers by removing cross-service dependencies and conflicts. Each container can be considered a different microservice and thus can be upgraded independently without fear of synchronization. 7. The runtime environment portable container encapsulates all relevant details necessary to run the application such as application dependencies and operating systems. This makes migration of images from one environment to another more flexible. For example, the same image may run in Windows or Linux or in a development, test or stage environment. 8. The established paradigm of an extensibility container allows users to extend applications with a relatively simple mechanism. The features of lightweight mirroring, fast boot time, creation of test and deployment of "golden images" and standardized runtime environments make it possible to build highly extended systems. 9. The processes between the independently upgradeable containers are isolated from each other, as are the infrastructure therein. Such that upgrades or changes to one of the containers do not affect the other containers.
In the prior art, the container has a plurality of advantages, but the current navigation service system cannot intelligently recommend the container of interest to the user according to the use interest of the user, so that the user can only use the whole navigation service system or manually select the container every time the user uses the navigation service system, the efficiency of the whole navigation service system is low, and the user has no good user experience, so that a technology is needed, and the user can only dynamically select the container with different software functions according to the preference of the user, thereby greatly improving the user experience and providing the efficiency of the navigation service system.
Disclosure of Invention
In order to solve the technical problems, the invention provides a container recommendation method of a navigation service system based on machine learning, which comprises the following steps:
acquiring historical data of a user using the navigation service system, and carrying out vectorization processing on the historical data to generate a historical data feature vector, wherein the historical data feature vector comprises: the user uses the frequency feature vector of each container, the time-long-in-use feature vector of each container, and the number of times-pruned feature vector of each container;
Extracting a feature vector of each container, wherein the feature vector comprises: a resource occupancy feature vector for each container, an operational status feature vector for each container, and a software functional feature vector provided by each container;
modeling and training the feature vector of each container through a machine learning model, classifying all containers to obtain container types of users, setting a precise container recommendation model according to the container types and the historical data feature vector, calculating first similarity between each container and the users, and recommending the container with the highest first similarity to the users;
and setting an inter-container similarity recommendation model, calculating the second similarity between the container with the highest first similarity and other containers in the container category of the container with the highest first similarity, and recommending the container with the highest second similarity to a user.
Further, the accurate container recommendation model is:
,
wherein n is n feature vectors,is the +.o of the characteristic vector of the historical data>The value of the characteristic vector is set,is the%>Individual characteristic vector values, < >>For the history data feature vector +.>The weights of the individual feature vectors are used, Is->No. 4 of individual containers>Number of occurrences of the individual feature vector,/->For the history data feature vector +.>The weight of each feature vector, I is the number of all containers.
Further, the similarity recommendation model between containers is as follows:
,
wherein ,is a container->In->Individual characteristic vector values, < >>Is a container->In->Individual characteristic vector values, < >>Is a container->Average value of all eigenvectors,/>Is a container->And (3) an average value of all the eigenvectors, wherein n is n eigenvectors.
Further, the method further comprises the following steps:
according to the accurate container recommendation model, calculating the third similarity in each container category, finding out the container with the highest third similarity in each container category, and recommending the container to a user;
and finding out the rest containers with the highest similarity with the third containers in each container category, and recommending the rest containers to the user.
Further, the method comprises the steps of:
and carrying out normalization processing on the historical data feature vector and the feature vector of each container to generate a feature vector value of the historical data feature vector and a feature vector value of the feature vector of each container.
The invention also provides a container recommendation system of the navigation service system based on machine learning, which comprises:
The navigation service system comprises an acquisition historical data module, a navigation service system and a navigation data processing module, wherein the acquisition historical data module is used for acquiring historical data of a user using the navigation service system, vectorizing the historical data and generating a historical data feature vector, and the historical data feature vector comprises: the user uses the frequency feature vector of each container, the time-long-in-use feature vector of each container, and the number of times-pruned feature vector of each container;
the device comprises a container feature vector obtaining module, a container feature vector extracting module and a container feature vector extracting module, wherein the feature vector comprises: a resource occupancy feature vector for each container, an operational status feature vector for each container, and a software functional feature vector provided by each container;
the accurate model module is used for modeling and training the feature vector of each container through the machine learning model, classifying all the containers to obtain container types of users, setting the accurate container recommendation model according to the container types and the historical data feature vector, calculating the first similarity between each container and the users, and recommending the container with the highest first similarity to the users;
and setting an inter-container model, namely setting an inter-container similarity recommendation model, calculating second similarity between the container with the highest first similarity and other containers in the container category of the container with the highest first similarity, and recommending the container with the highest second similarity to a user.
Further, the accurate container recommendation model is:
,
wherein n is n feature vectors,is the +.o of the characteristic vector of the historical data>The value of the characteristic vector is set,is the%>Individual characteristic vector values, < >>For the history data feature vector +.>The weights of the individual feature vectors are used,is->No. 4 of individual containers>Number of occurrences of the individual feature vector,/->For the history data feature vector +.>The weight of each feature vector, I is the number of all containers.
Further, the similarity recommendation model between containers is as follows:
,
wherein ,is a container->In->Individual characteristic vector values, < >>Is a container->In->Individual characteristic vector values, < >>Is a container->Average value of all eigenvectors,/>Is a container->And (3) an average value of all the eigenvectors, wherein n is n eigenvectors.
Further, the method further comprises the following steps:
according to the accurate container recommendation model, calculating the third similarity in each container category, finding out the container with the highest third similarity in each container category, and recommending the container to a user;
and finding out the rest containers with the highest similarity with the third containers in each container category, and recommending the rest containers to the user.
Further, the method comprises the steps of:
and carrying out normalization processing on the historical data feature vector and the feature vector of each container to generate a feature vector value of the historical data feature vector and a feature vector value of the feature vector of each container.
In general, the above technical solutions conceived by the present invention have the following beneficial effects compared with the prior art:
the invention obtains the historical data of the user using the navigation service system, and carries out vectorization processing on the historical data to generate a historical data feature vector, wherein the historical data feature vector comprises: the user uses the frequency feature vector of each container, the time-long-in-use feature vector of each container, and the number of times-pruned feature vector of each container; extracting a feature vector of each container, wherein the feature vector comprises: a resource occupancy feature vector for each container, an operational status feature vector for each container, and a software functional feature vector provided by each container; modeling and training the feature vector of each container through a machine learning model, classifying all containers to obtain container types of users, setting a precise container recommendation model according to the container types and the historical data feature vector, calculating first similarity between each container and the users, and recommending the container with the highest first similarity to the users; and setting an inter-container similarity recommendation model, calculating the second similarity between the container with the highest first similarity and other containers in the container category of the container with the highest first similarity, and recommending the container with the highest second similarity to a user. The containers, i.e., the functions of interest to the user, can be intelligently recommended according to the interests of the user.
Drawings
FIG. 1 is a flow chart of the method of embodiment 1 of the present invention;
FIG. 2 is a block diagram of the system of embodiment 2 of the present invention;
fig. 3 is a diagram of the technical architecture of the present invention.
Detailed Description
In order to better understand the above technical solutions, the following detailed description will be given with reference to the accompanying drawings and specific embodiments.
The method provided by the invention can be implemented in a terminal environment, wherein the terminal can comprise one or more of the following components: processor, storage medium, and display screen. Wherein the storage medium has stored therein at least one instruction that is loaded and executed by the processor to implement the method described in the embodiments below.
The processor may include one or more processing cores. The processor connects various parts within the overall terminal using various interfaces and lines, performs various functions of the terminal and processes data by executing or executing instructions, programs, code sets, or instruction sets stored in the storage medium, and invoking data stored in the storage medium.
The storage medium may include a random access Memory (Random Access Memory, RAM) or a Read-Only Memory (ROM). The storage medium may be used to store instructions, programs, code sets, or instructions.
The display screen is used for displaying a user interface of each application program.
In addition, it will be appreciated by those skilled in the art that the structure of the terminal described above is not limiting and that the terminal may include more or fewer components, or may combine certain components, or a different arrangement of components. For example, the terminal further includes components such as a radio frequency circuit, an input unit, a sensor, an audio circuit, a power supply, and the like, which are not described herein.
Aiming at the characteristics that the shipborne intelligent terminal has limited calculation and storage resources compared with the traditional computer terminal, the stability, the safety, the universality and the like of an operating system are poor, the application requirements of the terminal on high reliability and resource optimization configuration due to multi-task concurrency, multi-class peripheral random access and long-time communication positioning resource occupation which are required to be met under the actual working condition environment are considered, the navigation management and service integration application based on a container engine is provided, the universal container is constructed, the unified adaptation, unified deployment, unified upgrading, unified resource management and unified safety management and control of multi-class and multi-functional navigation management and service application based on elastic configuration access are realized, the terminal has sufficient service software expansion capability, and meanwhile, the decoupling of the service application and the terminal operating system is realized, and further, the technical foundation is provided for realizing the cross-mobile operating system platform application. The container can uniformly manage the 20 rest system applications, effectively reduces the utilization rate of the whole system resources, and reduces the running stability risks brought to other applications and operating systems by single application breakdown. The navigation management and service integration technical architecture based on the container engine is shown in fig. 3.
Example 1
As shown in fig. 1, an embodiment of the present invention provides a container recommendation method of a navigation service system based on machine learning, including:
step 101, obtaining historical data of a user using the navigation service system, and carrying out vectorization processing on the historical data to generate a historical data feature vector, wherein the historical data feature vector comprises: the user uses the frequency feature vector of each container, the time-long-in-use feature vector of each container, and the number of times-pruned feature vector of each container; normalizing the historical data feature vector and the feature vector of each container to generate a feature vector value of the historical data feature vector and a feature vector value of the feature vector of each container;
step 102, extracting a feature vector of each container, wherein the feature vector comprises: a resource occupancy feature vector for each container, an operational status feature vector for each container, and a software functional feature vector provided by each container;
step 103, modeling and training the feature vector of each container through a machine learning model, classifying all containers to obtain container types of users, setting a precise container recommendation model according to the container types and the historical data feature vector, calculating the first similarity between each container and the users, and recommending the container with the highest first similarity to the users; for example, feature vectors of containers are modeled and trained using machine learning models including, for example, clustering models or classification models. The clustering model can classify similar containers into different categories, and the classification model can classify each container into different categories, such as containers with good running stability, containers with less occupied resources, and the like, cluster or classify the containers used by the user according to the historical use records of the user, and obtain the container categories of the user.
Specifically, the accurate container recommendation model is:
,
wherein n is n feature vectors,is the +.o of the characteristic vector of the historical data>The value of the characteristic vector is set,is the%>Individual characteristic vector values, < >>For the history data feature vector +.>The weights of the individual feature vectors are used,is->No. 4 of individual containers>Number of occurrences of the individual feature vector,/->For the history data feature vector +.>The weight of each feature vector, I is the number of all containers.
And 104, setting a similarity recommendation model among the containers, calculating the second similarity between the container with the highest first similarity and the rest containers in the container category of the container with the highest first similarity, and recommending the container with the highest second similarity to a user.
Specifically, the similarity recommendation model between containers is as follows:
,
wherein ,is a container->In->Individual characteristic vector values, < >>Is a container->In->Individual characteristic vector values, < >>Is a container->Average value of all eigenvectors,/>Is a container->And (3) an average value of all the eigenvectors, wherein n is n eigenvectors.
Specifically, the method further comprises the following steps:
according to the accurate container recommendation model, calculating the third similarity in each container category, finding out the container with the highest third similarity in each container category, and recommending the container to a user;
And finding out the rest containers with the highest similarity with the third containers in each container category, and recommending the rest containers to the user.
The following is a specific example for describing modeling and training of feature vectors of each container by machine learning model in step 103, as follows:
the specific steps for training a decision tree model are as follows:
first, collecting data: a data set is collected with a label indicating whether the container has been used by the user, the data set should contain some characteristics, such as user use, characteristics of the container, etc.
Second, data are prepared: the data set is converted into a format suitable for processing by a decision tree algorithm, such as discretizing continuous features, processing missing values, and the like.
Thirdly, training a model: the data is trained using a decision tree algorithm and appropriate splitting criteria and stopping conditions are selected.
Fourth, test model: and verifying the model by using the test set, and evaluating performance indexes of the model, such as accuracy, recall, F1 score and the like.
Fifth, using a model: the trained model is used to predict new data, predicting the containers that the user is likely to use.
Specifically, the step of training the decision tree model is as follows:
Step 1, selecting splitting standard: depending on the different types of features (discrete or continuous), suitable splitting criteria are chosen, such as information gain, information gain ratio, keni purity, etc.
Step 2, selecting splitting attribute: the optimal splitting attribute is selected from the current node, so that the two split child nodes can distinguish the positive example and the negative example to the greatest extent.
Step 3, creating a child node: two child nodes are created according to the value of the splitting attribute, and the data set is divided into different child nodes according to the value of the splitting attribute.
Step 4, recursively constructing a decision tree: repeating the steps for each child node until the stopping condition is met, such as the maximum number of nodes, the minimum number of samples, the threshold value of node purity, and the like.
Step 5, pruning: pruning is carried out on the decision tree, and some decision nodes and leaf nodes are removed, so that overfitting is avoided, and the generalization capability of the model is improved.
The above is a specific step of training the decision tree model.
The specific steps of discretizing continuous features and processing missing values for the data set are as follows:
discretizing continuous features:
discretization of the continuous features may convert it into classification features for better processing by the decision tree algorithm. The discretization method comprises equal width discretization, equal frequency discretization, clustering discretization and the like.
Equal width discretization: the value range of the continuous feature is divided into a fixed number of intervals, each interval representing a discrete value, and the method is suitable for features with larger value ranges.
Equal frequency discretization: the values of the successive features are divided into a fixed number of intervals, each interval containing a substantially equal number of samples, suitable for features having a smaller range of values but a non-uniform distribution.
Clustering discretization: and clustering the values of the continuous features by using a clustering algorithm, and dividing each sample into discrete values corresponding to the nearest clustering center, thereby being suitable for the features with complex data distribution.
Processing the missing values: the missing values refer to the situation where the characteristic values of some samples are missing or unavailable. The processing of missing values can improve the stability and effect of decision tree algorithms, and common methods include deleting missing values, replacing missing values with mode, mean, median, etc., using interpolation methods, etc.
Deletion of missing values: if the number of missing values is not large, the samples where the missing values are located can be deleted, which has the disadvantage of reducing the sample size of the data set.
Instead of the missing value: the defect of deleting samples can be avoided by replacing the missing values with statistics such as mode, mean value, median and the like, but the distribution of the data set can be influenced to a certain extent.
Interpolation method: for continuous features, interpolation methods can be used to estimate missing values, such as linear interpolation, polynomial interpolation, spline interpolation, etc., and the distribution features of the dataset can be better maintained.
The following is another specific example for describing modeling and training of feature vectors of each container by machine learning model in step 103, as follows:
to solve this problem using machine learning algorithms, we need to prepare a labeled dataset containing information about the user's usage and container. For each user use, we can record the containers they use and the frequency and duration of use of the containers. These data may be used to train a machine learning model and to predict future containers that the user may use.
One common machine learning algorithm is a decision tree algorithm. The decision tree algorithm may divide the data set into a plurality of subsets, each subset having different features and labels. By selecting the most discriminating features, the decision tree can progressively separate the data sets and output labels at leaf nodes. In the case of container selection, we can train a decision tree model with the container as the label, and the user's use and other relevant features as input features. When new user usage inputs are available, the decision tree model can predict the containers that the user is likely to use based on these characteristics.
Example 2
As shown in fig. 2, the embodiment of the present invention further provides a container recommendation system of a navigation service system based on machine learning, including:
the navigation service system comprises an acquisition historical data module, a navigation service system and a navigation data processing module, wherein the acquisition historical data module is used for acquiring historical data of a user using the navigation service system, vectorizing the historical data and generating a historical data feature vector, and the historical data feature vector comprises: the user uses the frequency feature vector of each container, the time-long-in-use feature vector of each container, and the number of times-pruned feature vector of each container; normalizing the historical data feature vector and the feature vector of each container to generate a feature vector value of the historical data feature vector and a feature vector value of the feature vector of each container;
the device comprises a container feature vector obtaining module, a container feature vector extracting module and a container feature vector extracting module, wherein the feature vector comprises: a resource occupancy feature vector for each container, an operational status feature vector for each container, and a software functional feature vector provided by each container;
the accurate model module is used for modeling and training the feature vector of each container through the machine learning model, classifying all the containers to obtain container types of users, setting the accurate container recommendation model according to the container types and the historical data feature vector, calculating the first similarity between each container and the users, and recommending the container with the highest first similarity to the users; for example, feature vectors of containers are modeled and trained using machine learning models including, for example, clustering models or classification models. The clustering model can classify similar containers into different categories, and the classification model can classify each container into different categories, such as containers with good running stability, containers with less occupied resources, and the like, cluster or classify the containers used by the user according to the historical use records of the user, and obtain the container categories of the user.
Specifically, the accurate container recommendation model is:
,
wherein n is n feature vectors,is the +.o of the characteristic vector of the historical data>The value of the characteristic vector is set,is the%>Individual characteristic vector values, < >>For the history data feature vector +.>The weights of the individual feature vectors are used,is->No. 4 of individual containers>Number of occurrences of the individual feature vector,/->For the history data feature vector +.>Personal specialThe weight of the sign vector, I, is the number of all containers.
And setting an inter-container model, namely setting an inter-container similarity recommendation model, calculating second similarity between the container with the highest first similarity and other containers in the container category of the container with the highest first similarity, and recommending the container with the highest second similarity to a user.
Specifically, the similarity recommendation model between containers is as follows:
,
wherein ,is a container->In->Individual characteristic vector values, < >>Is a container->In->Individual characteristic vector values, < >>Is a container->Average value of all eigenvectors,/>Is a container->And (3) an average value of all the eigenvectors, wherein n is n eigenvectors.
Specifically, the method further comprises the following steps:
according to the accurate container recommendation model, calculating the third similarity in each container category, finding out the container with the highest third similarity in each container category, and recommending the container to a user;
And finding out the rest containers with the highest similarity with the third containers in each container category, and recommending the rest containers to the user.
The following is a specific example for describing modeling and training of feature vectors for each container by a machine learning model in the set-up precision model module, as follows:
the specific steps for training a decision tree model are as follows:
first, collecting data: a data set is collected with a label indicating whether the container has been used by the user, the data set should contain some characteristics, such as user use, characteristics of the container, etc.
Second, data are prepared: the data set is converted into a format suitable for processing by a decision tree algorithm, such as discretizing continuous features, processing missing values, and the like.
Thirdly, training a model: the data is trained using a decision tree algorithm and appropriate splitting criteria and stopping conditions are selected.
Fourth, test model: and verifying the model by using the test set, and evaluating performance indexes of the model, such as accuracy, recall, F1 score and the like.
Fifth, using a model: the trained model is used to predict new data, predicting the containers that the user is likely to use.
Specifically, the step of training the decision tree model is as follows:
step 1, selecting splitting standard: depending on the different types of features (discrete or continuous), suitable splitting criteria are chosen, such as information gain, information gain ratio, keni purity, etc.
Step 2, selecting splitting attribute: the optimal splitting attribute is selected from the current node, so that the two split child nodes can distinguish the positive example and the negative example to the greatest extent.
Step 3, creating a child node: two child nodes are created according to the value of the splitting attribute, and the data set is divided into different child nodes according to the value of the splitting attribute.
Step 4, recursively constructing a decision tree: repeating the steps for each child node until the stopping condition is met, such as the maximum number of nodes, the minimum number of samples, the threshold value of node purity, and the like.
Step 5, pruning: pruning is carried out on the decision tree, and some decision nodes and leaf nodes are removed, so that overfitting is avoided, and the generalization capability of the model is improved.
The above is a specific step of training the decision tree model.
The specific steps of discretizing continuous features and processing missing values for the data set are as follows:
discretizing continuous features:
Discretization of the continuous features may convert it into classification features for better processing by the decision tree algorithm. The discretization method comprises equal width discretization, equal frequency discretization, clustering discretization and the like.
Equal width discretization: the value range of the continuous feature is divided into a fixed number of intervals, each interval representing a discrete value, and the method is suitable for features with larger value ranges.
Equal frequency discretization: the values of the successive features are divided into a fixed number of intervals, each interval containing a substantially equal number of samples, suitable for features having a smaller range of values but a non-uniform distribution.
Clustering discretization: and clustering the values of the continuous features by using a clustering algorithm, and dividing each sample into discrete values corresponding to the nearest clustering center, thereby being suitable for the features with complex data distribution.
Processing the missing values: the missing values refer to the situation where the characteristic values of some samples are missing or unavailable. The processing of missing values can improve the stability and effect of decision tree algorithms, and common methods include deleting missing values, replacing missing values with mode, mean, median, etc., using interpolation methods, etc.
Deletion of missing values: if the number of missing values is not large, the samples where the missing values are located can be deleted, which has the disadvantage of reducing the sample size of the data set.
Instead of the missing value: the defect of deleting samples can be avoided by replacing the missing values with statistics such as mode, mean value, median and the like, but the distribution of the data set can be influenced to a certain extent.
Interpolation method: for continuous features, interpolation methods can be used to estimate missing values, such as linear interpolation, polynomial interpolation, spline interpolation, etc., and the distribution features of the dataset can be better maintained.
The following is another specific example for describing modeling and training of feature vectors of each container by machine learning model in the set-up precision model module, as follows:
to solve this problem using machine learning algorithms, we need to prepare a labeled dataset containing information about the user's usage and container. For each user use, we can record the containers they use and the frequency and duration of use of the containers. These data may be used to train a machine learning model and to predict future containers that the user may use.
One common machine learning algorithm is a decision tree algorithm. The decision tree algorithm may divide the data set into a plurality of subsets, each subset having different features and labels. By selecting the most discriminating features, the decision tree can progressively separate the data sets and output labels at leaf nodes. In the case of container selection, we can train a decision tree model with the container as the label, and the user's use and other relevant features as input features. When new user usage inputs are available, the decision tree model can predict the containers that the user is likely to use based on these characteristics.
Example 3
The embodiment of the invention also provides a storage medium which stores a plurality of instructions for realizing the container recommendation method of the navigation service system based on machine learning.
Alternatively, in this embodiment, the storage medium may be located in any one of the computer terminals in the computer terminal group in the computer network, or in any one of the mobile terminals in the mobile terminal group.
Alternatively, in the present embodiment, the storage medium is configured to store program code for performing the steps of: step 101, obtaining historical data of a user using the navigation service system, and carrying out vectorization processing on the historical data to generate a historical data feature vector, wherein the historical data feature vector comprises: the user uses the frequency feature vector of each container, the time-long-in-use feature vector of each container, and the number of times-pruned feature vector of each container; normalizing the historical data feature vector and the feature vector of each container to generate a feature vector value of the historical data feature vector and a feature vector value of the feature vector of each container;
Step 102, extracting a feature vector of each container, wherein the feature vector comprises: a resource occupancy feature vector for each container, an operational status feature vector for each container, and a software functional feature vector provided by each container;
step 103, modeling and training the feature vector of each container through a machine learning model, classifying all containers to obtain container types of users, setting a precise container recommendation model according to the container types and the historical data feature vector, calculating the first similarity between each container and the users, and recommending the container with the highest first similarity to the users; for example, feature vectors of containers are modeled and trained using machine learning models including, for example, clustering models or classification models. The clustering model can classify similar containers into different categories, and the classification model can classify each container into different categories, such as containers with good running stability, containers with less occupied resources, and the like, cluster or classify the containers used by the user according to the historical use records of the user, and obtain the container categories of the user.
Specifically, the accurate container recommendation model is:
,
Wherein n is n feature vectors,is the +.o of the characteristic vector of the historical data>The value of the characteristic vector is set,is the%>Individual characteristic vector values, < >>For the history data feature vector +.>The weights of the individual feature vectors are used,is->No. 4 of individual containers>Number of occurrences of the individual feature vector,/->For the history data feature vector +.>The weight of each feature vector, I is the number of all containers.
And 104, setting a similarity recommendation model among the containers, calculating the second similarity between the container with the highest first similarity and the rest containers in the container category of the container with the highest first similarity, and recommending the container with the highest second similarity to a user.
Specifically, the similarity recommendation model between containers is as follows:
,
wherein ,is a container->In->Individual characteristic vector values, < >>Is a container->In->Individual characteristic vector values, < >>Is a container->Average value of all eigenvectors,/>Is a container->And (3) an average value of all the eigenvectors, wherein n is n eigenvectors.
Specifically, the method further comprises the following steps:
according to the accurate container recommendation model, calculating the third similarity in each container category, finding out the container with the highest third similarity in each container category, and recommending the container to a user;
And finding out the rest containers with the highest similarity with the third containers in each container category, and recommending the rest containers to the user.
The following is a specific example for describing modeling and training of feature vectors of each container by machine learning model in step 103, as follows:
the specific steps for training a decision tree model are as follows:
first, collecting data: a data set is collected with a label indicating whether the container has been used by the user, the data set should contain some characteristics, such as user use, characteristics of the container, etc.
Second, data are prepared: the data set is converted into a format suitable for processing by a decision tree algorithm, such as discretizing continuous features, processing missing values, and the like.
Thirdly, training a model: the data is trained using a decision tree algorithm and appropriate splitting criteria and stopping conditions are selected.
Fourth, test model: and verifying the model by using the test set, and evaluating performance indexes of the model, such as accuracy, recall, F1 score and the like.
Fifth, using a model: the trained model is used to predict new data, predicting the containers that the user is likely to use.
Specifically, the step of training the decision tree model is as follows:
Step 1, selecting splitting standard: depending on the different types of features (discrete or continuous), suitable splitting criteria are chosen, such as information gain, information gain ratio, keni purity, etc.
Step 2, selecting splitting attribute: the optimal splitting attribute is selected from the current node, so that the two split child nodes can distinguish the positive example and the negative example to the greatest extent.
Step 3, creating a child node: two child nodes are created according to the value of the splitting attribute, and the data set is divided into different child nodes according to the value of the splitting attribute.
Step 4, recursively constructing a decision tree: repeating the steps for each child node until the stopping condition is met, such as the maximum number of nodes, the minimum number of samples, the threshold value of node purity, and the like.
Step 5, pruning: pruning is carried out on the decision tree, and some decision nodes and leaf nodes are removed, so that overfitting is avoided, and the generalization capability of the model is improved.
The above is a specific step of training the decision tree model.
The specific steps of discretizing continuous features and processing missing values for the data set are as follows:
discretizing continuous features:
discretization of the continuous features may convert it into classification features for better processing by the decision tree algorithm. The discretization method comprises equal width discretization, equal frequency discretization, clustering discretization and the like.
Equal width discretization: the value range of the continuous feature is divided into a fixed number of intervals, each interval representing a discrete value, and the method is suitable for features with larger value ranges.
Equal frequency discretization: the values of the successive features are divided into a fixed number of intervals, each interval containing a substantially equal number of samples, suitable for features having a smaller range of values but a non-uniform distribution.
Clustering discretization: and clustering the values of the continuous features by using a clustering algorithm, and dividing each sample into discrete values corresponding to the nearest clustering center, thereby being suitable for the features with complex data distribution.
Processing the missing values: the missing values refer to the situation where the characteristic values of some samples are missing or unavailable. The processing of missing values can improve the stability and effect of decision tree algorithms, and common methods include deleting missing values, replacing missing values with mode, mean, median, etc., using interpolation methods, etc.
Deletion of missing values: if the number of missing values is not large, the samples where the missing values are located can be deleted, which has the disadvantage of reducing the sample size of the data set.
Instead of the missing value: the defect of deleting samples can be avoided by replacing the missing values with statistics such as mode, mean value, median and the like, but the distribution of the data set can be influenced to a certain extent.
Interpolation method: for continuous features, interpolation methods can be used to estimate missing values, such as linear interpolation, polynomial interpolation, spline interpolation, etc., and the distribution features of the dataset can be better maintained.
The following is another specific example for describing modeling and training of feature vectors of each container by machine learning model in step 103, as follows:
to solve this problem using machine learning algorithms, we need to prepare a labeled dataset containing information about the user's usage and container. For each user use, we can record the containers they use and the frequency and duration of use of the containers. These data may be used to train a machine learning model and to predict future containers that the user may use.
One common machine learning algorithm is a decision tree algorithm. The decision tree algorithm may divide the data set into a plurality of subsets, each subset having different features and labels. By selecting the most discriminating features, the decision tree can progressively separate the data sets and output labels at leaf nodes. In the case of container selection, we can train a decision tree model with the container as the label, and the user's use and other relevant features as input features. When new user usage inputs are available, the decision tree model can predict the containers that the user is likely to use based on these characteristics.
Example 4
The embodiment of the invention also provides electronic equipment, which comprises a processor and a storage medium connected with the processor, wherein the storage medium stores a plurality of instructions, and the instructions can be loaded and executed by the processor so that the processor can execute the container recommendation method of the navigation service system based on machine learning.
Specifically, the electronic device of the present embodiment may be a computer terminal, and the computer terminal may include: one or more processors, and a storage medium.
The storage medium may be used to store a software program and a module, for example, in an embodiment of the present invention, a container recommendation method for a navigation service system based on machine learning, corresponding program instructions/modules, and the processor executes various function applications and data processing by running the software program and the module stored in the storage medium, thereby implementing the container recommendation method for the navigation service system based on machine learning. The storage medium may include a high-speed random access storage medium, and may also include a non-volatile storage medium, such as one or more magnetic storage systems, flash memory, or other non-volatile solid-state storage medium. In some examples, the storage medium may further include a storage medium remotely located with respect to the processor, and the remote storage medium may be connected to the terminal through a network. Examples of such networks include, but are not limited to, the internet, intranets, local area networks, mobile communication networks, and combinations thereof.
The processor may invoke the information stored in the storage medium and the application program via the transmission system to perform the following steps: step 101, obtaining historical data of a user using the navigation service system, and carrying out vectorization processing on the historical data to generate a historical data feature vector, wherein the historical data feature vector comprises: the user uses the frequency feature vector of each container, the time-long-in-use feature vector of each container, and the number of times-pruned feature vector of each container; normalizing the historical data feature vector and the feature vector of each container to generate a feature vector value of the historical data feature vector and a feature vector value of the feature vector of each container;
step 102, extracting a feature vector of each container, wherein the feature vector comprises: a resource occupancy feature vector for each container, an operational status feature vector for each container, and a software functional feature vector provided by each container;
step 103, modeling and training the feature vector of each container through a machine learning model, classifying all containers to obtain container types of users, setting a precise container recommendation model according to the container types and the historical data feature vector, calculating the first similarity between each container and the users, and recommending the container with the highest first similarity to the users; for example, feature vectors of containers are modeled and trained using machine learning models including, for example, clustering models or classification models. The clustering model can classify similar containers into different categories, and the classification model can classify each container into different categories, such as containers with good running stability, containers with less occupied resources, and the like, cluster or classify the containers used by the user according to the historical use records of the user, and obtain the container categories of the user.
Specifically, the accurate container recommendation model is:
,
wherein n is n feature vectors,is the +.o of the characteristic vector of the historical data>The value of the characteristic vector is set,is the%>Individual characteristic vector values, < >>For the history data feature vector +.>The weights of the individual feature vectors are used,is->No. 4 of individual containers>Number of occurrences of the individual feature vector,/->For the history data feature vector +.>The weight of each feature vector, I is the number of all containers.
And 104, setting a similarity recommendation model among the containers, calculating the second similarity between the container with the highest first similarity and the rest containers in the container category of the container with the highest first similarity, and recommending the container with the highest second similarity to a user.
Specifically, the similarity recommendation model between containers is as follows:
,
wherein ,is a container->In->Individual characteristic vector values, < >>Is a container->In->Individual characteristic vector values, < >>Is a container->All feature vectorsAverage value->Is a container->And (3) an average value of all the eigenvectors, wherein n is n eigenvectors.
Specifically, the method further comprises the following steps:
according to the accurate container recommendation model, calculating the third similarity in each container category, finding out the container with the highest third similarity in each container category, and recommending the container to a user;
And finding out the rest containers with the highest similarity with the third containers in each container category, and recommending the rest containers to the user.
The following is a specific example for describing modeling and training of feature vectors of each container by machine learning model in step 103, as follows:
the specific steps for training a decision tree model are as follows:
first, collecting data: a data set is collected with a label indicating whether the container has been used by the user, the data set should contain some characteristics, such as user use, characteristics of the container, etc.
Second, data are prepared: the data set is converted into a format suitable for processing by a decision tree algorithm, such as discretizing continuous features, processing missing values, and the like.
Thirdly, training a model: the data is trained using a decision tree algorithm and appropriate splitting criteria and stopping conditions are selected.
Fourth, test model: and verifying the model by using the test set, and evaluating performance indexes of the model, such as accuracy, recall, F1 score and the like.
Fifth, using a model: the trained model is used to predict new data, predicting the containers that the user is likely to use.
Specifically, the step of training the decision tree model is as follows:
Step 1, selecting splitting standard: depending on the different types of features (discrete or continuous), suitable splitting criteria are chosen, such as information gain, information gain ratio, keni purity, etc.
Step 2, selecting splitting attribute: the optimal splitting attribute is selected from the current node, so that the two split child nodes can distinguish the positive example and the negative example to the greatest extent.
Step 3, creating a child node: two child nodes are created according to the value of the splitting attribute, and the data set is divided into different child nodes according to the value of the splitting attribute.
Step 4, recursively constructing a decision tree: repeating the steps for each child node until the stopping condition is met, such as the maximum number of nodes, the minimum number of samples, the threshold value of node purity, and the like.
Step 5, pruning: pruning is carried out on the decision tree, and some decision nodes and leaf nodes are removed, so that overfitting is avoided, and the generalization capability of the model is improved.
The above is a specific step of training the decision tree model.
The specific steps of discretizing continuous features and processing missing values for the data set are as follows:
discretizing continuous features:
discretization of the continuous features may convert it into classification features for better processing by the decision tree algorithm. The discretization method comprises equal width discretization, equal frequency discretization, clustering discretization and the like.
Equal width discretization: the value range of the continuous feature is divided into a fixed number of intervals, each interval representing a discrete value, and the method is suitable for features with larger value ranges.
Equal frequency discretization: the values of the successive features are divided into a fixed number of intervals, each interval containing a substantially equal number of samples, suitable for features having a smaller range of values but a non-uniform distribution.
Clustering discretization: and clustering the values of the continuous features by using a clustering algorithm, and dividing each sample into discrete values corresponding to the nearest clustering center, thereby being suitable for the features with complex data distribution.
Processing the missing values: the missing values refer to the situation where the characteristic values of some samples are missing or unavailable. The processing of missing values can improve the stability and effect of decision tree algorithms, and common methods include deleting missing values, replacing missing values with mode, mean, median, etc., using interpolation methods, etc.
Deletion of missing values: if the number of missing values is not large, the samples where the missing values are located can be deleted, which has the disadvantage of reducing the sample size of the data set.
Instead of the missing value: the defect of deleting samples can be avoided by replacing the missing values with statistics such as mode, mean value, median and the like, but the distribution of the data set can be influenced to a certain extent.
Interpolation method: for continuous features, interpolation methods can be used to estimate missing values, such as linear interpolation, polynomial interpolation, spline interpolation, etc., and the distribution features of the dataset can be better maintained.
The following is another specific example for describing modeling and training of feature vectors of each container by machine learning model in step 103, as follows:
to solve this problem using machine learning algorithms, we need to prepare a labeled dataset containing information about the user's usage and container. For each user use, we can record the containers they use and the frequency and duration of use of the containers. These data may be used to train a machine learning model and to predict future containers that the user may use.
One common machine learning algorithm is a decision tree algorithm. The decision tree algorithm may divide the data set into a plurality of subsets, each subset having different features and labels. By selecting the most discriminating features, the decision tree can progressively separate the data sets and output labels at leaf nodes. In the case of container selection, we can train a decision tree model with the container as the label, and the user's use and other relevant features as input features. When new user usage inputs are available, the decision tree model can predict the containers that the user is likely to use based on these characteristics.
The foregoing embodiment numbers of the present invention are merely for the purpose of description, and do not represent the advantages or disadvantages of the embodiments.
In the foregoing embodiments of the present invention, the descriptions of the embodiments are emphasized, and for a portion of this disclosure that is not described in detail in this embodiment, reference is made to the related descriptions of other embodiments.
In the embodiments provided in the present invention, it should be understood that the disclosed technology may be implemented in other manners. The system embodiments described above are merely exemplary, and for example, the division of the units is merely a logic function division, and there may be another division manner in actual implementation, for example, multiple units or components may be combined or integrated into another system, or some features may be omitted or not performed. Alternatively, the coupling or direct coupling or communication connection shown or discussed with each other may be through some interfaces, units or modules, or may be in electrical or other forms.
The units described as separate units may or may not be physically separate, and units shown as units may or may not be physical units, may be located in one place, or may be distributed on a plurality of network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.
In addition, each functional unit in the embodiments of the present invention may be integrated in one processing unit, or each unit may exist alone physically, or two or more units may be integrated in one unit. The integrated units may be implemented in hardware or in software functional units.
The integrated units, if implemented in the form of software functional units and sold or used as stand-alone products, may be stored in a computer readable storage medium. Based on such understanding, the technical solution of the present invention may be embodied essentially or partly in the form of a software product or all or part of the technical solution, which is stored in a storage medium, and includes several instructions for causing a computer device (which may be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: a U-disk, a Read-Only Memory (ROM), a random access Memory (RAM, randomAccess Memory), a removable hard disk, a magnetic disk, or an optical disk, or the like, which can store program codes.
It is apparent that the above examples are given by way of illustration only and are not limiting of the embodiments. Other variations or modifications of the above teachings will be apparent to those of ordinary skill in the art. It is not necessary here nor is it exhaustive of all embodiments. While still being apparent from variations or modifications that may be made by those skilled in the art are within the scope of the invention.
Claims (10)
1. A container recommendation method for a navigation service system based on machine learning, comprising:
acquiring historical data of a user using the navigation service system, and carrying out vectorization processing on the historical data to generate a historical data feature vector, wherein the historical data feature vector comprises: the user uses the frequency feature vector of each container, the time-long-in-use feature vector of each container, and the number of times-pruned feature vector of each container;
extracting a feature vector of each container, wherein the feature vector comprises: a resource occupancy feature vector for each container, an operational status feature vector for each container, and a software functional feature vector provided by each container;
Modeling and training the feature vector of each container through a machine learning model, classifying all containers to obtain container types of users, setting a precise container recommendation model according to the container types and the historical data feature vector, calculating first similarity between each container and the users, and recommending the container with the highest first similarity to the users;
and setting an inter-container similarity recommendation model, calculating the second similarity between the container with the highest first similarity and other containers in the container category of the container with the highest first similarity, and recommending the container with the highest second similarity to a user.
2. The method for recommending containers for a navigation service system based on machine learning according to claim 1, wherein the accurate container recommendation model is as follows:
,
wherein n is n feature vectors,is the +.o of the characteristic vector of the historical data>Individual characteristic vector values, < >>Is the%>Individual characteristic vector values, < >>For the history data feature vector +.>Weights of individual feature vectors, +.>Is the firstFirst container of/>Number of occurrences of the individual feature vector,/->For the history data feature vector +. >Weights of individual feature vectors, +.>For all containers.
3. The method for recommending containers for a navigation service system based on machine learning according to claim 2, wherein the model for recommending similarity between containers is as follows:
,
wherein ,is a container->In->Individual characteristic vector values, < >>Is a container->In->Individual characteristic vector values, < >>Is a container->Average value of all eigenvectors,/>Is a container->And (3) an average value of all the eigenvectors, wherein n is n eigenvectors.
4. The machine learning based navigation service system container recommendation method of claim 1, further comprising:
according to the accurate container recommendation model, calculating the third similarity in each container category, finding out the container with the highest third similarity in each container category, and recommending the container to a user;
and finding out the rest containers with the highest similarity with the third containers in each container category, and recommending the rest containers to the user.
5. The machine learning based navigation service system container recommendation method of claim 1, comprising:
and carrying out normalization processing on the historical data feature vector and the feature vector of each container to generate a feature vector value of the historical data feature vector and a feature vector value of the feature vector of each container.
6. A container recommendation system for a machine learning based navigation service system, comprising:
the navigation service system comprises an acquisition historical data module, a navigation service system and a navigation data processing module, wherein the acquisition historical data module is used for acquiring historical data of a user using the navigation service system, vectorizing the historical data and generating a historical data feature vector, and the historical data feature vector comprises: the user uses the frequency feature vector of each container, the time-long-in-use feature vector of each container, and the number of times-pruned feature vector of each container;
the device comprises a container feature vector obtaining module, a container feature vector extracting module and a container feature vector extracting module, wherein the feature vector comprises: a resource occupancy feature vector for each container, an operational status feature vector for each container, and a software functional feature vector provided by each container;
the accurate model module is used for modeling and training the feature vector of each container through the machine learning model, classifying all the containers to obtain container types of users, setting the accurate container recommendation model according to the container types and the historical data feature vector, calculating the first similarity between each container and the users, and recommending the container with the highest first similarity to the users;
And setting an inter-container model, namely setting an inter-container similarity recommendation model, calculating second similarity between the container with the highest first similarity and other containers in the container category of the container with the highest first similarity, and recommending the container with the highest second similarity to a user.
7. The machine learning based navigation service system of claim 6, wherein the accurate container recommendation model is:
,
wherein n is n feature vectors,is the +.o of the characteristic vector of the historical data>Individual characteristic vector values, < >>Is the%>Individual characteristic vector values, < >>For the history data feature vector +.>Weights of individual feature vectors, +.>Is the firstNo. 4 of individual containers>Number of occurrences of the individual feature vector,/->For the history data feature vector +.>The weight of each feature vector, I is the number of all containers.
8. The machine learning based navigation service system of claim 7, wherein the inter-container similarity recommendation model is:
,
wherein ,is a container->In->Individual characteristic vector values, < >>Is a container->In->Individual characteristic vector values, < > >Is a container->Average value of all eigenvectors,/>Is a container->And (3) an average value of all the eigenvectors, wherein n is n eigenvectors.
9. The machine learning based navigation service system container recommendation system of claim 6, further comprising:
according to the accurate container recommendation model, calculating the third similarity in each container category, finding out the container with the highest third similarity in each container category, and recommending the container to a user;
and finding out the rest containers with the highest similarity with the third containers in each container category, and recommending the rest containers to the user.
10. The machine learning based navigation service system container recommendation system of claim 6, comprising:
and carrying out normalization processing on the historical data feature vector and the feature vector of each container to generate a feature vector value of the historical data feature vector and a feature vector value of the feature vector of each container.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310767471.4A CN116522002B (en) | 2023-06-27 | 2023-06-27 | Container recommendation method and system of navigation service system based on machine learning |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202310767471.4A CN116522002B (en) | 2023-06-27 | 2023-06-27 | Container recommendation method and system of navigation service system based on machine learning |
Publications (2)
Publication Number | Publication Date |
---|---|
CN116522002A true CN116522002A (en) | 2023-08-01 |
CN116522002B CN116522002B (en) | 2023-09-08 |
Family
ID=87408513
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202310767471.4A Active CN116522002B (en) | 2023-06-27 | 2023-06-27 | Container recommendation method and system of navigation service system based on machine learning |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN116522002B (en) |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107908457A (en) * | 2017-11-08 | 2018-04-13 | 河海大学 | A kind of containerization cloud resource distribution method based on stable matching |
CN113094116A (en) * | 2021-04-01 | 2021-07-09 | 中国科学院软件研究所 | Deep learning application cloud configuration recommendation method and system based on load characteristic analysis |
CN114168252A (en) * | 2020-08-20 | 2022-03-11 | 中国电信股份有限公司 | Information processing system and method, network scheme recommendation component and method |
US11314630B1 (en) * | 2020-12-14 | 2022-04-26 | International Business Machines Corporation | Container configuration recommendations |
CN115022098A (en) * | 2022-08-09 | 2022-09-06 | 北京瑞莱智慧科技有限公司 | Artificial intelligence safety target range content recommendation method, device and storage medium |
CN116107991A (en) * | 2021-11-10 | 2023-05-12 | 中国电信股份有限公司 | Container label database construction method and device, storage medium and electronic equipment |
-
2023
- 2023-06-27 CN CN202310767471.4A patent/CN116522002B/en active Active
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107908457A (en) * | 2017-11-08 | 2018-04-13 | 河海大学 | A kind of containerization cloud resource distribution method based on stable matching |
CN114168252A (en) * | 2020-08-20 | 2022-03-11 | 中国电信股份有限公司 | Information processing system and method, network scheme recommendation component and method |
US11314630B1 (en) * | 2020-12-14 | 2022-04-26 | International Business Machines Corporation | Container configuration recommendations |
CN113094116A (en) * | 2021-04-01 | 2021-07-09 | 中国科学院软件研究所 | Deep learning application cloud configuration recommendation method and system based on load characteristic analysis |
CN116107991A (en) * | 2021-11-10 | 2023-05-12 | 中国电信股份有限公司 | Container label database construction method and device, storage medium and electronic equipment |
CN115022098A (en) * | 2022-08-09 | 2022-09-06 | 北京瑞莱智慧科技有限公司 | Artificial intelligence safety target range content recommendation method, device and storage medium |
Non-Patent Citations (1)
Title |
---|
施超;谢在鹏;柳晗;吕鑫;: "基于稳定匹配的容器部署策略的优化", 计算机科学, no. 04 * |
Also Published As
Publication number | Publication date |
---|---|
CN116522002B (en) | 2023-09-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN108628947B (en) | Business rule matching processing method, device and processing equipment | |
US11640563B2 (en) | Automated data processing and machine learning model generation | |
CN108804641B (en) | Text similarity calculation method, device, equipment and storage medium | |
CN111507768B (en) | Potential user determination method and related device | |
CN112163625B (en) | Big data mining method based on artificial intelligence and cloud computing and cloud service center | |
CN112800095B (en) | Data processing method, device, equipment and storage medium | |
CN109948710B (en) | Micro-service identification method based on API similarity | |
CN108415845A (en) | AB tests computational methods, device and the server of system index confidence interval | |
CN110414569B (en) | Clustering implementation method and device | |
EP4198771A1 (en) | Data processing method and apparatus, computer readable medium, and electronic device | |
CN110708285B (en) | Flow monitoring method, device, medium and electronic equipment | |
CN113626241B (en) | Abnormality processing method, device, equipment and storage medium for application program | |
CN110909222A (en) | User portrait establishing method, device, medium and electronic equipment based on clustering | |
CN112115162A (en) | Big data processing method based on e-commerce cloud computing and artificial intelligence server | |
CN116663938B (en) | Informatization management method based on enterprise data center system and related device thereof | |
CN111507400A (en) | Application classification method and device, electronic equipment and storage medium | |
CN113159213A (en) | Service distribution method, device and equipment | |
CN116522002B (en) | Container recommendation method and system of navigation service system based on machine learning | |
CN113761017A (en) | Similarity searching method and device | |
CN112114907A (en) | Application loading method based on e-commerce cloud computing and artificial intelligence computing center | |
CN116932147A (en) | Streaming job processing method and device, electronic equipment and medium | |
CN112948251B (en) | Automatic software testing method and device | |
CN111723872B (en) | Pedestrian attribute identification method and device, storage medium and electronic device | |
CN114385918A (en) | Message pushing method and device, computer equipment and storage medium | |
CN113435343B (en) | Image recognition method and device, computer equipment and storage medium |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
GR01 | Patent grant | ||
GR01 | Patent grant |