CN111898766A - Ether house fuel limitation prediction method and device based on automatic machine learning - Google Patents
Ether house fuel limitation prediction method and device based on automatic machine learning Download PDFInfo
- Publication number
- CN111898766A CN111898766A CN202010761121.3A CN202010761121A CN111898766A CN 111898766 A CN111898766 A CN 111898766A CN 202010761121 A CN202010761121 A CN 202010761121A CN 111898766 A CN111898766 A CN 111898766A
- Authority
- CN
- China
- Prior art keywords
- intelligent contract
- machine learning
- target
- contract code
- automatic machine
- 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
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
- G06N20/00—Machine learning
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/20—Information retrieval; Database structures therefor; File system structures therefor of structured data, e.g. relational data
- G06F16/27—Replication, distribution or synchronisation of data between databases or within a distributed database system; Distributed database system architectures therefor
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/213—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods
- G06F18/2135—Feature extraction, e.g. by transforming the feature space; Summarisation; Mappings, e.g. subspace methods based on approximation criteria, e.g. principal component analysis
-
- 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
-
- 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
- G06N3/08—Learning methods
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
- G06Q40/04—Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange
Landscapes
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Computation (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Life Sciences & Earth Sciences (AREA)
- Software Systems (AREA)
- Computing Systems (AREA)
- Bioinformatics & Computational Biology (AREA)
- Evolutionary Biology (AREA)
- Mathematical Physics (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Business, Economics & Management (AREA)
- Molecular Biology (AREA)
- Finance (AREA)
- Biomedical Technology (AREA)
- Biophysics (AREA)
- Computational Linguistics (AREA)
- General Health & Medical Sciences (AREA)
- Databases & Information Systems (AREA)
- Health & Medical Sciences (AREA)
- Accounting & Taxation (AREA)
- Medical Informatics (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Marketing (AREA)
- Strategic Management (AREA)
- Technology Law (AREA)
- General Business, Economics & Management (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses an Ethernet workshop fuel limit prediction method based on automatic machine learning, a device, computer equipment and a storage medium, which relate to a block chain technology and comprise the steps of acquiring network addresses of all intelligent contracts issued on an Ethernet workshop from a target website so as to acquire a target intelligent contract code set which is verified and transaction information corresponding to each target intelligent contract code; inputting the corresponding characteristic set after information screening of each transaction information into an automatic machine learning model to be trained for training to obtain the automatic machine learning model; and if the current intelligent contract code is detected, acquiring a current feature set corresponding to the current intelligent contract code and inputting the current feature set into the automatic machine learning model for operation to obtain the corresponding Ether house fuel limit. After automatic feature dimension reduction based on intelligent contract code automatic screening is realized, the Ethengfang fuel limitation is predicted, manual intervention is avoided, labor cost is reduced, and prediction accuracy is improved.
Description
Technical Field
The invention relates to the technical field of block chains, in particular to an Ethenhouse fuel limit prediction method and device based on automatic machine learning, computer equipment and a storage medium.
Background
The most core innovation of the successful item-bitcoin in the blockchain is that the value can be transferred from a long distance without trusting a third party. However, bitcoin has the disadvantage that the pictographic script language is not supported. That is, the bitcoin is only stored under the condition of the distributed environment, but is not both stored and calculated under the distributed condition. To address this problem, Vitalik et al introduced EtherFang. Compared with bitcoin, the biggest difference of Ethenfang is that Ethenfang can support scripting language with complete graphic and allows developers to develop any application on the Ethenfang to realize intelligent contracts.
The etherhouse implements a runtime environment on the blockchain, called the etherhouse virtual machine. Each node participating in the ethernet lane network will run an ethernet lane virtual machine as part of the block verification protocol. These nodes will validate each transaction that is covered in the block and run the transaction triggered code (code inside the smart contracts) in the ethernet virtual machine. All nodes on each network would perform the same calculations and store the same values. While each command, such as add, hash, etc., has a specific consumption in executing the codes and calculations, counting with fuel at the ether house, for example, adding at the ether house would consume 3 fuels.
Since certain fuel is consumed in the code execution process, the fuel consumption is also related to the state of the intelligent contract. The user therefore pays in advance a certain amount of fuel before each transaction is made. This prepaid amount is simply referred to as a fuel limit in the ether house. During authentication and calculation at nodes on the network, if the user's transaction is used to calculate that the amount of fuel needed to be used is less than or equal to the set fuel limit, then the transaction is processed. Conversely, if the total consumption of fuel exceeds the fuel limit, the fuel supplied by the user is used up, and even all operations are resumed during the process. Therefore, it is very important to ensure the accuracy of the fuel limitation.
Machine learning, a branch of artificial intelligence, is also a popular research topic. The machine learning algorithm is an algorithm for automatically analyzing and obtaining rules from data and predicting unknown data by using the rules. The greatest advantage of machine learning is that the working efficiency is greatly improved. Machine learning does not solve human unsolved problems, but it can accept large amounts of data, quickly establish connections based on the data, and make predictions. Therefore, it is efficient and accurate to use machine learning to make predictions when large amounts of data are collected. And tens of millions of transactions are already available in the current ether house. Finding the regularity in these data using machine learning is clearly a good approach. Therefore, it is feasible to predict the fuel limit traded on the ether house using machine learning. However, the data of the EtherFang trading program contains the intelligent contract URL, the characteristics are not obvious, the workload of manual characteristic engineering and manual selection of a machine learning model is large, and the universality and the accuracy are difficult to ensure.
Disclosure of Invention
The embodiment of the invention provides an Ethenhouse fuel limit prediction method and device based on automatic machine learning, computer equipment and a storage medium, and aims to solve the problems that in the prior art, Ethenhouse transaction program data comprise network addresses of intelligent contracts, data characteristics are not obvious, manual characteristic engineering and manual selection of a machine learning model have large workload, and universality and accuracy are difficult to guarantee.
In a first aspect, an embodiment of the present invention provides an ethernet fuel limit prediction method based on automatic machine learning, which includes:
calling a preset stored breadth-first algorithm and a preset target website, and acquiring network addresses of all intelligent contracts published on an Ethernet workshop from the target website through breadth-first search corresponding to the breadth-first algorithm;
acquiring a target intelligent contract code set which is verified and transaction information corresponding to each target intelligent contract code in the target intelligent contract code set according to the network address;
calling a pre-stored information field screening strategy, and carrying out information screening on the transaction information corresponding to each target intelligent contract code to obtain a feature set corresponding to each target intelligent contract code; the information field screening strategy is used for screening core features in the transaction information corresponding to the intelligent contract codes to form a feature set;
acquiring a feature set corresponding to each target intelligent contract code, inputting the feature set to an automatic machine learning model to be trained, and training to obtain the automatic machine learning model; wherein the automatic machine learning model is to predict a fuel limit of a function called by an intelligent contract;
if the current intelligent contract code uploaded by the user side is detected, acquiring a current feature set corresponding to the current intelligent contract code according to the information field screening strategy; and
and inputting the current feature set into the automatic machine learning model for operation to obtain the Ether house fuel limit corresponding to the current intelligent contract code, and sending the Ether house fuel limit corresponding to the current intelligent contract code to a corresponding target receiving end.
In a second aspect, an embodiment of the present invention provides an automatic machine learning-based ether house fuel limit prediction apparatus, which includes:
the target network address acquisition unit is used for calling a preset stored breadth-first algorithm and a preset target website, and acquiring network addresses of all intelligent contracts published on the Ethernet from the target website through breadth-first search corresponding to the breadth-first algorithm;
a target code set acquiring unit, configured to acquire, according to the network address, a target intelligent contract code set for which verification has been completed and transaction information corresponding to each target intelligent contract code in the target intelligent contract code set;
the feature set acquisition unit is used for calling a pre-stored information field screening strategy, and after the transaction information corresponding to each target intelligent contract code is subjected to information screening, the feature set corresponding to each target intelligent contract code is obtained; the information field screening strategy is used for screening core features in the transaction information corresponding to the intelligent contract codes to form a feature set;
the automatic machine learning model training unit is used for acquiring feature sets corresponding to target intelligent contract codes and inputting the feature sets to the automatic machine learning model to be trained for training to obtain an automatic machine learning model; wherein the automatic machine learning model is to predict a fuel limit of a function called by an intelligent contract;
the current feature set acquisition unit is used for acquiring a current feature set corresponding to a current intelligent contract code according to the information field screening strategy if the current intelligent contract code uploaded by the user side is detected; and
and the fuel limit prediction unit is used for inputting the current feature set into the automatic machine learning model for operation to obtain the Ether house fuel limit corresponding to the current intelligent contract code, and sending the Ether house fuel limit corresponding to the current intelligent contract code to a corresponding target receiving end.
In a third aspect, an embodiment of the present invention further provides a computer device, which includes a memory, a processor, and a computer program stored in the memory and executable on the processor, and the processor, when executing the computer program, implements the automatic machine learning-based etherhouse fuel limitation prediction method according to the first aspect.
In a fourth aspect, the present invention further provides a computer-readable storage medium, where the computer-readable storage medium stores a computer program, and the computer program, when executed by a processor, causes the processor to execute the method for predicting the etherhouse fuel limitation based on automatic machine learning according to the first aspect.
The embodiment of the invention provides an Ethernet workshop fuel limitation prediction method, a device, computer equipment and a storage medium based on automatic machine learning, which comprises the steps of obtaining network addresses of all intelligent contracts published on an Ethernet workshop from a target website through breadth-first search corresponding to a breadth-first algorithm; acquiring a target intelligent contract code set which is verified and transaction information corresponding to each target intelligent contract code in the target intelligent contract code set according to the network address; calling an information field screening strategy, and carrying out information screening on the transaction information corresponding to each target intelligent contract code to obtain a feature set corresponding to each target intelligent contract code; acquiring a feature set corresponding to each target intelligent contract code, inputting the feature set to an automatic machine learning model to be trained, and training to obtain the automatic machine learning model; if the current intelligent contract code uploaded by the user side is detected, obtaining a current feature set corresponding to the current intelligent contract code according to an information field screening strategy; and inputting the current feature set into an automatic machine learning model for operation to obtain the Ether house fuel limit corresponding to the current intelligent contract code, and sending the Ether house fuel limit corresponding to the current intelligent contract code to the corresponding target receiving end. After automatic feature dimension reduction based on intelligent contract code automatic screening is realized, the Ethengfang fuel limitation is predicted, manual intervention is avoided, labor cost is reduced, and prediction accuracy is improved.
Drawings
In order to more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings needed to be used in the description of the embodiments are briefly introduced below, and it is obvious that the drawings in the following description are some embodiments of the present invention, and it is obvious for those skilled in the art to obtain other drawings based on these drawings without creative efforts.
Fig. 1 is a schematic view of an application scenario of an ether house fuel limit prediction method based on automatic machine learning according to an embodiment of the present invention;
FIG. 2 is a schematic flow chart of a method for predicting Etherhouse fuel limit based on automatic machine learning according to an embodiment of the present invention;
FIG. 3 is a schematic block diagram of an automatic machine learning-based Etherhouse fuel limit prediction apparatus provided in accordance with an embodiment of the present invention;
FIG. 4 is a schematic block diagram of a computer device provided by an embodiment of the present invention.
Detailed Description
The technical solutions in the embodiments of the present invention will be clearly and completely described below with reference to the drawings in the embodiments of the present invention, and it is obvious that the described embodiments are some, not all, embodiments of the present invention. All other embodiments, which can be derived by a person skilled in the art from the embodiments given herein without making any creative effort, shall fall within the protection scope of the present invention.
It will be understood that the terms "comprises" and/or "comprising," when used in this specification and the appended claims, specify the presence of stated features, integers, steps, operations, elements, and/or components, but do not preclude the presence or addition of one or more other features, integers, steps, operations, elements, components, and/or groups thereof.
It is also to be understood that the terminology used in the description of the invention herein is for the purpose of describing particular embodiments only and is not intended to be limiting of the invention. As used in the specification of the present invention and the appended claims, the singular forms "a," "an," and "the" are intended to include the plural forms as well, unless the context clearly indicates otherwise.
It should be further understood that the term "and/or" as used in this specification and the appended claims refers to and includes any and all possible combinations of one or more of the associated listed items.
Referring to fig. 1 and fig. 2, fig. 1 is a schematic view of an application scenario of an automatic machine learning-based ether house fuel limit prediction method according to an embodiment of the present invention; fig. 2 is a schematic flowchart of an automatic machine learning-based ether house fuel limit prediction method according to an embodiment of the present invention, where the automatic machine learning-based ether house fuel limit prediction method is applied to a server, and the method is executed by application software installed in the server.
As shown in fig. 2, the method includes steps S110 to S160.
S110, calling a preset stored breadth first algorithm and a preset target website, and acquiring network addresses of all intelligent contracts published on the Ethernet by breadth first search corresponding to the breadth first algorithm from the target website.
In this embodiment, verified intelligent contracts on the ethernet archways and transaction information related to the intelligent contracts are stored on target websites (the target websites are specifically http:// etherscan. io /), and network addresses where all the intelligent contracts are published on the ethernet archways can be collected from all levels of webpage contents of the target websites through breadth-first search corresponding to a breadth-first algorithm. Through the breadth-first searching mode, the network addresses of all intelligent contracts published on the Etherns are obtained in a traversing mode, and the obtained network address set data volume is richer.
In one embodiment, step S110 includes:
acquiring all network addresses of all intelligent contracts issued on an ether house in a first-level webpage of the target website to form a first-level network address set;
accessing all second-level webpages adjacent to the first-level webpages, and acquiring all network addresses of all intelligent contracts issued on the ether houses in the second-level webpages to form a second-level network address set; sequentially accessing all third-level webpages adjacent to the second-level webpages until all nth-level webpages adjacent to the (n-1) th-level webpages are accessed so as to respectively obtain a third-level network address set to an nth-level network address set; wherein the value of n is equal to the total webpage level number of the target website;
and the network addresses of all intelligent contracts issued on the Ethernet workshops in the target website are formed by the first-level network address set to the nth-level network address set.
In the present embodiment, the breadth-first search algorithm (also called breadth-first search) is one of the simplest graph search algorithms, and this algorithm is also the prototype of many important graph algorithms. The core idea of the breadth-first algorithm is as follows: starting from an initial node, generating first-layer nodes by applying an operator, checking whether target nodes are in the subsequent nodes, if not, expanding all the first-layer nodes one by using a production rule to obtain second-layer nodes, and checking whether the second-layer nodes contain the target nodes one by one. If not, all nodes … … of the second level are expanded one by one with operators, and so on, and examined until a target node is found. The breadth-first algorithm is used, the searching depth is small, each node only needs to visit once, and the node is always visited in the shortest path, so that the efficiency of obtaining the network addresses of all intelligent contracts issued on an Ethernet workshop is improved.
And S120, acquiring the verified target intelligent contract code set and the transaction information corresponding to each target intelligent contract code in the target intelligent contract code set according to the network address.
In this embodiment, the verified intelligent contract on the target website is tagged with the verified label, and at this time, the intelligent contract code with the verified label can be quickly screened out to form the target intelligent contract code set. The transaction information corresponding to each intelligent contract has a plurality of field values, all the field values may be related to fuel limitation prediction, or part of the field values may be related to fuel limitation prediction, and core fields in the transaction information corresponding to the intelligent contract need to be screened in the subsequent steps.
In one embodiment, step S120 includes:
naming and storing each target intelligent contract code in the target intelligent contract code set of which the network address acquisition is verified;
acquiring the block height of the transaction, the hash value of the transaction, the fuel limit, the fuel actually used for executing the transaction independently and the input data of the function used by the transaction, which are included in the transaction information of each target intelligent contract code; the input data of the function used by the exchange comprises the number of times of executing the SHA256 function by the exchange, the number of times of executing the SHA3 function by the exchange, the number of FOR loops in the function executed by the exchange and the number of variables in the exchange.
In this embodiment, since some intelligent contracts have several versions on the blockchain, the version names are all the same but the code of each version may not be the same. The intelligent contracts are named in a mode of intelligent contract name + intelligent contract version number + intelligent contract uploading time, and therefore intelligent contracts with the same version name but different codes are distinguished through different names.
The method for acquiring the transaction information of the intelligent contract mainly comprises the following steps: the height of the block where the transaction is located, the hash value of the transaction, the fuel limit, the fuel actually used for executing the transaction independently, and the input data of the function used by the exchange in the exchange; the input data of the function used by the exchange comprises the number of times of executing the SHA256 function by the exchange, the number of times of executing the SHA3 function by the exchange, the number of FOR loops in the function executed by the exchange and the number of variables in the exchange. According to the analysis of the data importance, the input data of the function used by the general exchange has the largest relevance to the fuel limit prediction, so that an information field screening strategy for screening the core field value in the target intelligent contract code can be set, and the follow-up practicability is facilitated.
S130, calling a pre-stored information field screening strategy, and carrying out information screening on the transaction information corresponding to each target intelligent contract code to obtain a feature set corresponding to each target intelligent contract code; the information field screening strategy is used for screening core features in the transaction information corresponding to the intelligent contract codes to form a feature set.
In this embodiment, because the transaction information corresponding to each target intelligent contract code includes a large number of fields, in order to reduce data dimensionality, the transaction information corresponding to each target intelligent contract code may be subjected to information screening to obtain a feature set corresponding to each target intelligent contract code.
In one embodiment, step S130 includes:
acquiring a core characteristic field set included in the information field screening strategy; the core characteristic field set comprises a block height field where a transaction is located, a SHA256 function time field executed by a transaction exchange, a SHA3 function time field executed by the transaction, an FOR cycle time field in a function executed by the transaction exchange and a number field of variables in the transaction;
and carrying out information screening on the transaction information corresponding to each target intelligent contract code according to the core characteristic field set to obtain a characteristic set corresponding to each target intelligent contract code.
In this embodiment, in the ethernet virtual machine environment, each operation consumes a part of the fuel, for example, the addition operation on the ethernet needs to consume 3 fuels. The above examples are fuel consumed by a single operation, and the fuel consumed by the function involved in the intelligent contract code is not the sum of the fuel consumed by the data operation. However, it is known that the core feature field included in the information field screening strategy corresponds to a value, which is in a positive correlation with fuel consumption.
In specific implementation, the core characteristic fields positively correlated with the fuel consumption comprise a block height field where a transaction is located, a SHA256 function number field of an exchange, a SHA3 function number field of the exchange, an FOR cycle number field of the exchange execution function and a number field of variables in the transaction, and the information field screening strategy is set as a corresponding value FOR screening the core characteristic fields included in the characteristic set. And correspondingly carrying out field screening on the transaction information corresponding to each target intelligent contract code, wherein the obtained screening result is the feature set corresponding to each target intelligent contract code. Through the screening process, the dimension of the data features is effectively reduced.
S140, acquiring a feature set corresponding to each target intelligent contract code, inputting the feature set to an automatic machine learning model to be trained, and training to obtain the automatic machine learning model; wherein the automatic machine learning model is to predict a fuel limit of a function called by the smart contract.
In this embodiment, machine learning is to let an algorithm automatically find out a set of rules from data, so as to extract features in the data that are helpful for classification/clustering/decision, as machine learning develops, there are more and more parts that need intervention manually, and AutoML (i.e. automatic machine learning) is to automate the whole process from building to applying of a machine learning model, and finally, an end-to-end model (end to end) is obtained.
The application of machine learning requires a large amount of manual intervention, which is manifested in: and (3) various aspects of machine learning such as feature extraction, model selection and parameter adjustment. Automatic machine learning (AutoML) attempts to automatically learn these important steps related to features, models, optimization, evaluation, so that the machine learning model can be applied without human intervention.
For a transaction fuel prediction task, an input data structure is complex, the input data structure also comprises characteristics such as code texts and the like which are difficult to quantify, the text cannot be directly accepted as the characteristics by using a traditional machine learning regression method, the difficulty is high when feature engineering and model selection are carried out, and the problem difficulty can be greatly reduced if an automatic machine learning method is used.
In one embodiment, step S140 includes:
calling a pre-stored principal component analysis algorithm to perform principal feature selection on feature sets corresponding to the target intelligent contract codes to obtain dimension reduction feature sets corresponding to the feature sets;
and sequentially carrying out model training, model selection/combination and hyper-parameter tuning on the automatic machine learning model to be trained according to the dimensionality reduction characteristic set to obtain the automatic machine learning model.
In the present embodiment, the principal component analysis algorithm, PCA, is mainly used for data dimension reduction. In the process of training the automatic machine learning model, the data dimensionality of the training set does not need to be too much, and at the moment, the main characteristic selection can be carried out on the characteristic set corresponding to each target intelligent contract code to obtain the dimensionality reduction characteristic set corresponding to each characteristic set, so that the data dimensionality reduction is realized.
And then, sequentially carrying out model training, model selection/combination and hyper-parameter tuning on the automatic machine learning model to be trained according to the dimension reduction characteristic set to obtain the automatic machine learning model.
The method comprises the steps of dividing from the flow sequence of automatic machine learning, firstly preparing data, including data collection and cleaning, and then performing feature engineering, wherein the feature engineering comprises feature selection, feature extraction (dimension reduction on features, common methods such as PCA) and feature combination (combining/constructing a plurality of features into a new feature); in the subsequent model construction, the most key is model selection, then hyper-parameter optimization can be carried out, a plurality of modes can be adopted, the simplest method is grid search, and the common method comprises the steps of using reinforcement learning, evolutionary algorithm, Bayesian optimization and gradient descent to reduce the search space; and finally, the AutoML introduces the process of stopping in advance, reducing the precision of the model and sharing parameters to automate the evaluation of the model.
The task of data collection is not to search and collect real data, but also to generate simulation data to expand the training data set, and new techniques that can be used include an anti-neural network and a reinforcement learning framework to optimize parameters for controlling the generated data, so that the generated data can be more effectively used for training the model. And the data cleaning is the work which is manually completed before missing value completion, outlier processing, characteristic normalization, different codes of the class type characteristics and the like are automatically completed.
The model is selected automatically, the traditional method is to select one or more models with the best effect by combining the traditional models, such as KNN, SVM and decision tree, and the current research focus of AutoML is Neural Architecture Search (english is called Neural Architecture Search), that is, the model automatically generates a network structure which is most effective for the current task without manual intervention.
The most common parameter adjusting process after model selection is grid search, that is, dotting is performed on a search space according to a fixed distance, and the AutoML uses random sampling, firstly evaluates the importance of each hyper-parameter, and then finely adjusts the important parameters.
And after the automatic machine learning model to be trained is trained through the dimension reduction characteristic set, the automatic machine learning model for predicting the fuel limit of the function called by the intelligent contract can be obtained.
S150, if the current intelligent contract code uploaded by the user side is detected, the current feature set corresponding to the current intelligent contract code is obtained according to the information field screening strategy.
In this embodiment, in order to predict the ether house fuel limit corresponding to other intelligent contract codes, the user may directly perform operation and entry of the current intelligent contract code on the user side, and then the user side sends the current intelligent contract code to the server, so as to perform data preprocessing of fuel limit prediction in the server. In this case, in step S130, the current feature set corresponding to the current intelligent contract code may also be obtained according to the information field screening policy. The data dimensionality is effectively reduced by predicting through the reduced current feature set, so that the subsequent prediction efficiency is improved.
And S160, inputting the current feature set into the automatic machine learning model for operation to obtain the Ether house fuel limit corresponding to the current intelligent contract code, and sending the Ether house fuel limit corresponding to the current intelligent contract code to a corresponding target receiving end.
In this embodiment, the current feature set is input into the automatic machine learning model for operation, and the fuel limit consumed by each function or data operation in the current feature set can be extracted through the automatic machine learning model, so that the ether house fuel limit corresponding to the current intelligent contract code is accurately predicted. At this time, in order to timely notify the user side of the prediction result, the ether house fuel limit corresponding to the current intelligent contract code may be sent to the corresponding target receiving end.
In one embodiment, step S160 includes:
performing main feature selection on the current feature set according to the main component analysis algorithm to obtain a current dimension reduction feature set corresponding to the current feature set;
inputting the current dimension reduction characteristic set into a model selection model in the automatic machine learning model for operation to obtain a target model;
and inputting the current feature set into the target model for operation to obtain the Ether house fuel limit corresponding to the current intelligent contract code.
In the embodiment, the fuel limitation prediction is performed through automatic machine learning, which mainly realizes the automatic selection of models, the traditional method is to select one or more models with the best combination effect from the traditional models, and the current AutoML can automatically generate a network structure which is most effective for the current task through neural architecture search, that is, without manual intervention, so that the accuracy of the prediction result is greatly improved.
The method realizes automatic dimension reduction based on the automatic feature screening of the intelligent contract codes so as to predict the Etheng fuel limitation, thereby not only avoiding manual intervention and reducing the labor cost, but also improving the accuracy of prediction.
The embodiment of the invention also provides an ether house fuel limit prediction device based on automatic machine learning, which is used for executing any embodiment of the ether house fuel limit prediction method based on automatic machine learning. Specifically, referring to fig. 3, fig. 3 is a schematic block diagram of an automatic machine learning-based ether house fuel limit prediction apparatus according to an embodiment of the present invention. The automated machine learning based ether house fuel limit prediction apparatus 100 may be configured in a server.
As shown in fig. 3, the automatic machine learning-based ether house fuel limit prediction apparatus 100 includes: a target network address acquisition unit 110, a target code set acquisition unit 120, a feature set acquisition unit 130, an automatic machine learning model training unit 140, a current feature set acquisition unit 150, and a fuel limit prediction unit 160.
And the target network address obtaining unit 110 is configured to invoke a preset stored breadth-first algorithm and a preset target website, and obtain network addresses of all intelligent contracts published on the ethernet archways from the target website through breadth-first search corresponding to the breadth-first algorithm.
In this embodiment, verified intelligent contracts on the ethernet archways and transaction information related to the intelligent contracts are stored on target websites (the target websites are specifically http:// etherscan. io /), and network addresses where all the intelligent contracts are published on the ethernet archways can be collected from all levels of webpage contents of the target websites through breadth-first search corresponding to a breadth-first algorithm. Through the breadth-first searching mode, the network addresses of all intelligent contracts published on the Etherns are obtained in a traversing mode, and the obtained network address set data volume is richer.
In one embodiment, the target network address obtaining unit 110 includes:
the first-level acquisition unit is used for acquiring all network addresses of all intelligent contracts issued on the ether house in a first-level webpage of the target website so as to form a first-level network address set;
traversing the next-level obtaining unit, which is used for accessing all second-level webpages adjacent to the first-level webpages and obtaining all network addresses of all intelligent contracts issued on the ether houses in the second-level webpages to form a second-level network address set; sequentially accessing all third-level webpages adjacent to the second-level webpages until all nth-level webpages adjacent to the (n-1) th-level webpages are accessed so as to respectively obtain a third-level network address set to an nth-level network address set; wherein the value of n is equal to the total webpage level number of the target website;
and the network address combination unit is used for forming network addresses of all intelligent contracts issued on the Etherns in the target website by the first-level network address set to the nth-level network address set.
In the present embodiment, the breadth-first search algorithm (also called breadth-first search) is one of the simplest graph search algorithms, and this algorithm is also the prototype of many important graph algorithms. The core idea of the breadth-first algorithm is as follows: starting from an initial node, generating first-layer nodes by applying an operator, checking whether target nodes are in the subsequent nodes, if not, expanding all the first-layer nodes one by using a production rule to obtain second-layer nodes, and checking whether the second-layer nodes contain the target nodes one by one. If not, all nodes … … of the second level are expanded one by one with operators, and so on, and examined until a target node is found. The breadth-first algorithm is used, the searching depth is small, each node only needs to visit once, and the node is always visited in the shortest path, so that the efficiency of obtaining the network addresses of all intelligent contracts issued on an Ethernet workshop is improved.
And a target code set obtaining unit 120, configured to obtain, according to the network address, a target intelligent contract code set for which verification is completed and transaction information corresponding to each target intelligent contract code in the target intelligent contract code set.
In this embodiment, the verified intelligent contract on the target website is tagged with the verified label, and at this time, the intelligent contract code with the verified label can be quickly screened out to form the target intelligent contract code set. The transaction information corresponding to each intelligent contract has a plurality of field values, all the field values may be related to fuel limitation prediction, or part of the field values may be related to fuel limitation prediction, and core fields in the transaction information corresponding to the intelligent contract need to be screened in the subsequent steps.
In one embodiment, the object code set obtaining unit 120 includes:
a naming storage unit, configured to name and store each target intelligent contract code in the target intelligent contract code set for which verification has been completed by the network address acquisition;
the transaction information analysis and acquisition unit is used for acquiring the block height of the transaction, the hash value of the transaction, the fuel limit, the fuel actually used for executing the transaction independently and the input data of the function used by the transaction, which are included in the transaction information of each target intelligent contract code; the input data of the function used by the exchange comprises the number of times of executing the SHA256 function by the exchange, the number of times of executing the SHA3 function by the exchange, the number of FOR loops in the function executed by the exchange and the number of variables in the exchange.
In this embodiment, since some intelligent contracts have several versions on the blockchain, the version names are all the same but the code of each version may not be the same. The intelligent contracts are named in a mode of intelligent contract name + intelligent contract version number + intelligent contract uploading time, and therefore intelligent contracts with the same version name but different codes are distinguished through different names.
The method for acquiring the transaction information of the intelligent contract mainly comprises the following steps: the height of the block where the transaction is located, the hash value of the transaction, the fuel limit, the fuel actually used for executing the transaction independently, and the input data of the function used by the exchange in the exchange; the input data of the function used by the exchange comprises the number of times of executing the SHA256 function by the exchange, the number of times of executing the SHA3 function by the exchange, the number of FOR loops in the function executed by the exchange and the number of variables in the exchange. According to the analysis of the data importance, the input data of the function used by the general exchange has the largest relevance to the fuel limit prediction, so that an information field screening strategy for screening the core field value in the target intelligent contract code can be set, and the follow-up practicability is facilitated.
The feature set obtaining unit 130 is configured to invoke a pre-stored information field screening policy, perform information screening on the transaction information corresponding to each target intelligent contract code, and obtain a feature set corresponding to each target intelligent contract code; the information field screening strategy is used for screening core features in the transaction information corresponding to the intelligent contract codes to form a feature set.
In this embodiment, because the transaction information corresponding to each target intelligent contract code includes a large number of fields, in order to reduce data dimensionality, the transaction information corresponding to each target intelligent contract code may be subjected to information screening to obtain a feature set corresponding to each target intelligent contract code.
In one embodiment, the feature set obtaining unit 130 includes:
an information field screening strategy obtaining unit, configured to obtain a core feature field set included in the information field screening strategy; the core characteristic field set comprises a block height field where a transaction is located, a SHA256 function time field executed by a transaction exchange, a SHA3 function time field executed by the transaction, an FOR cycle time field in a function executed by the transaction exchange and a number field of variables in the transaction;
and the core characteristic field acquisition unit is used for carrying out information screening on the transaction information corresponding to each target intelligent contract code according to the core characteristic field set to obtain a characteristic set corresponding to each target intelligent contract code.
In this embodiment, in the ethernet virtual machine environment, each operation consumes a part of the fuel, for example, the addition operation on the ethernet needs to consume 3 fuels. The above examples are fuel consumed by a single operation, and the fuel consumed by the function involved in the intelligent contract code is not the sum of the fuel consumed by the data operation. However, it is known that the core feature field included in the information field screening strategy corresponds to a value, which is in a positive correlation with fuel consumption.
In specific implementation, the core characteristic fields positively correlated with the fuel consumption comprise a block height field where a transaction is located, a SHA256 function number field of an exchange, a SHA3 function number field of the exchange, an FOR cycle number field of the exchange execution function and a number field of variables in the transaction, and the information field screening strategy is set as a corresponding value FOR screening the core characteristic fields included in the characteristic set. And correspondingly carrying out field screening on the transaction information corresponding to each target intelligent contract code, wherein the obtained screening result is the feature set corresponding to each target intelligent contract code. Through the screening process, the dimension of the data features is effectively reduced.
The automatic machine learning model training unit 140 is configured to acquire feature sets corresponding to the target intelligent contract codes and input the feature sets to the automatic machine learning model to be trained for training, so as to obtain an automatic machine learning model; wherein the automatic machine learning model is to predict a fuel limit of a function called by the smart contract.
In this embodiment, machine learning is to let an algorithm automatically find out a set of rules from data, so as to extract features in the data that are helpful for classification/clustering/decision, as machine learning develops, there are more and more parts that need intervention manually, and AutoML (i.e. automatic machine learning) is to automate the whole process from building to applying of a machine learning model, and finally, an end-to-end model (end to end) is obtained.
The application of machine learning requires a large amount of manual intervention, which is manifested in: and (3) various aspects of machine learning such as feature extraction, model selection and parameter adjustment. Automatic machine learning (AutoML) attempts to automatically learn these important steps related to features, models, optimization, evaluation, so that the machine learning model can be applied without human intervention.
For a transaction fuel prediction task, an input data structure is complex, the input data structure also comprises characteristics such as code texts and the like which are difficult to quantify, the text cannot be directly accepted as the characteristics by using a traditional machine learning regression method, the difficulty is high when feature engineering and model selection are carried out, and the problem difficulty can be greatly reduced if an automatic machine learning method is used.
In an embodiment, the automatic machine learning model training unit 140 comprises:
the data dimension reduction unit is used for calling a pre-stored principal component analysis algorithm to select the principal features of the feature set corresponding to each target intelligent contract code to obtain a dimension reduction feature set corresponding to each feature set;
and the model training unit is used for sequentially carrying out model training, model selection/combination and hyper-parameter tuning on the automatic machine learning model to be trained according to the dimensionality reduction characteristic set to obtain the automatic machine learning model.
In the present embodiment, the principal component analysis algorithm, PCA, is mainly used for data dimension reduction. In the process of training the automatic machine learning model, the data dimensionality of the training set does not need to be too much, and at the moment, the main characteristic selection can be carried out on the characteristic set corresponding to each target intelligent contract code to obtain the dimensionality reduction characteristic set corresponding to each characteristic set, so that the data dimensionality reduction is realized.
And then, sequentially carrying out model training, model selection/combination and hyper-parameter tuning on the automatic machine learning model to be trained according to the dimension reduction characteristic set to obtain the automatic machine learning model.
The method comprises the steps of dividing from the flow sequence of automatic machine learning, firstly preparing data, including data collection and cleaning, and then performing feature engineering, wherein the feature engineering comprises feature selection, feature extraction (dimension reduction on features, common methods such as PCA) and feature combination (combining/constructing a plurality of features into a new feature); in the subsequent model construction, the most key is model selection, then hyper-parameter optimization can be carried out, a plurality of modes can be adopted, the simplest method is grid search, and the common method comprises the steps of using reinforcement learning, evolutionary algorithm, Bayesian optimization and gradient descent to reduce the search space; and finally, the AutoML introduces the process of stopping in advance, reducing the precision of the model and sharing parameters to automate the evaluation of the model.
The task of data collection is not to search and collect real data, but also to generate simulation data to expand the training data set, and new techniques that can be used include an anti-neural network and a reinforcement learning framework to optimize parameters for controlling the generated data, so that the generated data can be more effectively used for training the model. And the data cleaning is the work which is manually completed before missing value completion, outlier processing, characteristic normalization, different codes of the class type characteristics and the like are automatically completed.
The model is selected automatically, the traditional method is to select one or more models with the best effect by combining the traditional models, such as KNN, SVM and decision tree, and the current research focus of AutoML is Neural Architecture Search (english is called Neural Architecture Search), that is, the model automatically generates a network structure which is most effective for the current task without manual intervention.
The most common parameter adjusting process after model selection is grid search, that is, dotting is performed on a search space according to a fixed distance, and the AutoML uses random sampling, firstly evaluates the importance of each hyper-parameter, and then finely adjusts the important parameters.
And after the automatic machine learning model to be trained is trained through the dimension reduction characteristic set, the automatic machine learning model for predicting the fuel limit of the function called by the intelligent contract can be obtained.
And a current feature set obtaining unit 150, configured to, if a current intelligent contract code uploaded by the user side is detected, obtain, according to the information field screening policy, a current feature set corresponding to the current intelligent contract code.
In this embodiment, in order to predict the ether house fuel limit corresponding to other intelligent contract codes, the user may directly perform operation and entry of the current intelligent contract code on the user side, and then the user side sends the current intelligent contract code to the server, so as to perform data preprocessing of fuel limit prediction in the server. In this case, the feature set obtaining unit 130 may be also referred to, and obtain the current feature set corresponding to the current intelligent contract code according to the information field screening policy. The data dimensionality is effectively reduced by predicting through the reduced current feature set, so that the subsequent prediction efficiency is improved.
And the fuel limit prediction unit 160 is configured to input the current feature set into the automatic machine learning model for operation, obtain the ether house fuel limit corresponding to the current intelligent contract code, and send the ether house fuel limit corresponding to the current intelligent contract code to a corresponding target receiving end.
In this embodiment, the current feature set is input into the automatic machine learning model for operation, and the fuel limit consumed by each function or data operation in the current feature set can be extracted through the automatic machine learning model, so that the ether house fuel limit corresponding to the current intelligent contract code is accurately predicted. At this time, in order to timely notify the user side of the prediction result, the ether house fuel limit corresponding to the current intelligent contract code may be sent to the corresponding target receiving end.
In one embodiment, the fuel limit prediction unit 160 includes:
the current dimension reduction unit is used for selecting the main features of the current feature set according to the main component analysis algorithm and acquiring the current dimension reduction feature set corresponding to the current feature set;
the target model selection unit is used for inputting the current dimension reduction characteristic set into a model selection model in the automatic machine learning model for operation to obtain a target model;
and the current prediction unit is used for inputting the current feature set into the target model for operation to obtain the Ether house fuel limit corresponding to the current intelligent contract code.
In the embodiment, the fuel limitation prediction is performed through automatic machine learning, which mainly realizes the automatic selection of models, the traditional method is to select one or more models with the best combination effect from the traditional models, and the current AutoML can automatically generate a network structure which is most effective for the current task through neural architecture search, that is, without manual intervention, so that the accuracy of the prediction result is greatly improved.
The device realizes automatic dimension reduction based on intelligent contract code automatic screening characteristics to predict the ether workshop fuel limitation, thereby not only avoiding manual intervention and reducing labor cost, but also improving the accuracy of prediction.
The above-described automatic machine learning-based ether house fuel limit prediction apparatus may be implemented in the form of a computer program that is executable on a computer device as shown in fig. 4.
Referring to fig. 4, fig. 4 is a schematic block diagram of a computer device according to an embodiment of the present invention. The computer device 500 is a server, and the server may be an independent server or a server cluster composed of a plurality of servers.
Referring to fig. 4, the computer device 500 includes a processor 502, memory, and a network interface 505 connected by a system bus 501, where the memory may include a non-volatile storage medium 503 and an internal memory 504.
The non-volatile storage medium 503 may store an operating system 5031 and a computer program 5032. The computer program 5032, when executed, may cause the processor 502 to perform an automated machine learning-based ether house fuel limit prediction method.
The processor 502 is used to provide computing and control capabilities that support the operation of the overall computer device 500.
The internal memory 504 provides an environment for the execution of the computer program 5032 in the non-volatile storage medium 503, and when executed by the processor 502, the computer program 5032 causes the processor 502 to perform an automatic machine learning based etherhouse fuel limit prediction method.
The network interface 505 is used for network communication, such as providing transmission of data information. Those skilled in the art will appreciate that the configuration shown in fig. 4 is a block diagram of only a portion of the configuration associated with aspects of the present invention and is not intended to limit the computing device 500 to which aspects of the present invention may be applied, and that a particular computing device 500 may include more or less components than those shown, or may combine certain components, or have a different arrangement of components.
The processor 502 is configured to run a computer program 5032 stored in the memory to implement the automatic machine learning-based ether house fuel limit prediction method disclosed in the embodiment of the present invention.
Those skilled in the art will appreciate that the embodiment of a computer device illustrated in fig. 4 does not constitute a limitation on the specific construction of the computer device, and that in other embodiments a computer device may include more or fewer components than those illustrated, or some components may be combined, or a different arrangement of components. For example, in some embodiments, the computer device may only include a memory and a processor, and in such embodiments, the structures and functions of the memory and the processor are consistent with those of the embodiment shown in fig. 4, and are not described herein again.
It should be understood that, in the embodiment of the present invention, the Processor 502 may be a Central Processing Unit (CPU), and the Processor 502 may also be other general purpose processors, Digital Signal Processors (DSPs), Application Specific Integrated Circuits (ASICs), Field Programmable Gate Arrays (FPGAs) or other Programmable logic devices, discrete Gate or transistor logic devices, discrete hardware components, and the like. Wherein a general purpose processor may be a microprocessor or the processor may be any conventional processor or the like.
In another embodiment of the invention, a computer-readable storage medium is provided. The computer readable storage medium may be a non-volatile computer readable storage medium. The computer readable storage medium stores a computer program, wherein the computer program, when executed by a processor, implements the automatic machine learning-based ether house fuel limit prediction method disclosed by the embodiments of the present invention.
It is clear to those skilled in the art that, for convenience and brevity of description, the specific working processes of the above-described apparatuses, devices and units may refer to the corresponding processes in the foregoing method embodiments, and are not described herein again. Those of ordinary skill in the art will appreciate that the elements and algorithm steps of the examples described in connection with the embodiments disclosed herein may be embodied in electronic hardware, computer software, or combinations of both, and that the components and steps of the examples have been described in a functional general in the foregoing description for the purpose of illustrating clearly the interchangeability of hardware and software. Whether such functionality is implemented as hardware or software depends upon the particular application and design constraints imposed on the implementation. Skilled artisans may implement the described functionality in varying ways for each particular application, but such implementation decisions should not be interpreted as causing a departure from the scope of the present invention.
In the embodiments provided by the present invention, it should be understood that the disclosed apparatus, device and method can be implemented in other ways. For example, the above-described embodiments of the apparatus are merely illustrative, and for example, the division of the units is only a logical division, and there may be other divisions when the actual implementation is performed, or units having the same function may be grouped into one unit, for example, a plurality of units or components may be combined or may be integrated into another system, or some features may be omitted, or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection may be an indirect coupling or communication connection through some interfaces, devices or units, and may also be an electric, mechanical or other form of connection.
The units described as separate parts may or may not be physically separate, and parts displayed 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 can be selected according to actual needs to achieve the purpose of the solution of the embodiment of the present invention.
In addition, functional units in the embodiments of the present invention may be integrated into one processing unit, or each unit may exist alone physically, or two or more units are integrated into one unit. The integrated unit can be realized in a form of hardware, and can also be realized in a form of a software functional unit.
The integrated unit, if implemented in the form of a software functional unit and sold or used as a stand-alone product, may be stored in a storage medium. Based on such understanding, the technical solution of the present invention essentially or partially contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product stored in a storage medium and including instructions for causing a computer device (which may be a personal computer, a server, or a network device) to execute all or part of the steps of the method according to the embodiments of the present invention. And the aforementioned storage medium includes: various media capable of storing program codes, such as a usb disk, a removable hard disk, a Read-Only Memory (ROM), a magnetic disk, or an optical disk.
While the invention has been described with reference to specific embodiments, the invention is not limited thereto, and various equivalent modifications and substitutions can be easily made by those skilled in the art within the technical scope of the invention. Therefore, the protection scope of the present invention shall be subject to the protection scope of the claims.
Claims (10)
1. An Etherhouse fuel limit prediction method based on automatic machine learning, comprising:
calling a preset stored breadth-first algorithm and a preset target website, and acquiring network addresses of all intelligent contracts published on an Ethernet workshop from the target website through breadth-first search corresponding to the breadth-first algorithm;
acquiring a target intelligent contract code set which is verified and transaction information corresponding to each target intelligent contract code in the target intelligent contract code set according to the network address;
calling a pre-stored information field screening strategy, and carrying out information screening on the transaction information corresponding to each target intelligent contract code to obtain a feature set corresponding to each target intelligent contract code; the information field screening strategy is used for screening core features in the transaction information corresponding to the intelligent contract codes to form a feature set;
acquiring a feature set corresponding to each target intelligent contract code, inputting the feature set to an automatic machine learning model to be trained, and training to obtain the automatic machine learning model; wherein the automatic machine learning model is to predict a fuel limit of a function called by an intelligent contract;
if the current intelligent contract code uploaded by the user side is detected, acquiring a current feature set corresponding to the current intelligent contract code according to the information field screening strategy; and
and inputting the current feature set into the automatic machine learning model for operation to obtain the Ether house fuel limit corresponding to the current intelligent contract code, and sending the Ether house fuel limit corresponding to the current intelligent contract code to a corresponding target receiving end.
2. The automatic machine learning-based ether house fuel limit prediction method according to claim 1, wherein the network addresses of all intelligent contracts released on the ether house are acquired from the target web address through breadth-first search corresponding to a breadth-first algorithm, including;
acquiring all network addresses of all intelligent contracts issued on an ether house in a first-level webpage of the target website to form a first-level network address set;
accessing all second-level webpages adjacent to the first-level webpages, and acquiring all network addresses of all intelligent contracts issued on the ether houses in the second-level webpages to form a second-level network address set; sequentially accessing all third-level webpages adjacent to the second-level webpages until all nth-level webpages adjacent to the (n-1) th-level webpages are accessed so as to respectively obtain a third-level network address set to an nth-level network address set; wherein the value of n is equal to the total webpage level number of the target website;
and the network addresses of all intelligent contracts issued on the Ethernet workshops in the target website are formed by the first-level network address set to the nth-level network address set.
3. The automated machine learning-based etherhouse fuel limit prediction method of claim 1, wherein obtaining the validated set of target intelligent contract codes and the transaction information corresponding to each target intelligent contract code in the set of target intelligent contract codes based on the network address comprises:
naming and storing each target intelligent contract code in the target intelligent contract code set of which the network address acquisition is verified;
acquiring the block height of the transaction, the hash value of the transaction, the fuel limit, the fuel actually used for executing the transaction independently and the input data of the function used by the transaction, which are included in the transaction information of each target intelligent contract code; the input data of the function used by the exchange comprises the number of times of executing the SHA256 function by the exchange, the number of times of executing the SHA3 function by the exchange, the number of FOR loops in the function executed by the exchange and the number of variables in the exchange.
4. The automatic machine learning-based ether house fuel limit prediction method according to claim 3, wherein the invoking of a pre-stored information field screening policy to perform information screening on the transaction information corresponding to each target intelligent contract code to obtain a feature set corresponding to each target intelligent contract code comprises:
acquiring a core characteristic field set included in the information field screening strategy; the core characteristic field set comprises a block height field where a transaction is located, a SHA256 function time field executed by a transaction exchange, a SHA3 function time field executed by the transaction, an FOR cycle time field in a function executed by the transaction exchange and a number field of variables in the transaction;
and carrying out information screening on the transaction information corresponding to each target intelligent contract code according to the core characteristic field set to obtain a characteristic set corresponding to each target intelligent contract code.
5. The Etherhouse fuel limit prediction method based on automatic machine learning of claim 4, wherein the obtaining of the feature set corresponding to each target intelligent contract code is input to an automatic machine learning model to be trained for training to obtain the automatic machine learning model, comprising:
calling a pre-stored principal component analysis algorithm to perform principal feature selection on feature sets corresponding to the target intelligent contract codes to obtain dimension reduction feature sets corresponding to the feature sets;
and sequentially carrying out model training, model selection/combination and hyper-parameter tuning on the automatic machine learning model to be trained according to the dimensionality reduction characteristic set to obtain the automatic machine learning model.
6. The Etherhouse fuel limit prediction method based on automatic machine learning of claim 5, wherein said inputting the current feature set into the automatic machine learning model for computation to obtain the Etherhouse fuel limit corresponding to the current intelligent contract code comprises:
performing main feature selection on the current feature set according to the main component analysis algorithm to obtain a current dimension reduction feature set corresponding to the current feature set;
inputting the current dimension reduction characteristic set into a model selection model in the automatic machine learning model for operation to obtain a target model;
and inputting the current feature set into the target model for operation to obtain the Ether house fuel limit corresponding to the current intelligent contract code.
7. An ether house fuel limit prediction device based on automatic machine learning, comprising:
the target network address acquisition unit is used for calling a preset stored breadth-first algorithm and a preset target website, and acquiring network addresses of all intelligent contracts published on the Ethernet from the target website through breadth-first search corresponding to the breadth-first algorithm;
a target code set acquiring unit, configured to acquire, according to the network address, a target intelligent contract code set for which verification has been completed and transaction information corresponding to each target intelligent contract code in the target intelligent contract code set;
the feature set acquisition unit is used for calling a pre-stored information field screening strategy, and after the transaction information corresponding to each target intelligent contract code is subjected to information screening, the feature set corresponding to each target intelligent contract code is obtained; the information field screening strategy is used for screening core features in the transaction information corresponding to the intelligent contract codes to form a feature set;
the automatic machine learning model training unit is used for acquiring feature sets corresponding to target intelligent contract codes and inputting the feature sets to the automatic machine learning model to be trained for training to obtain an automatic machine learning model; wherein the automatic machine learning model is to predict a fuel limit of a function called by an intelligent contract;
the current feature set acquisition unit is used for acquiring a current feature set corresponding to a current intelligent contract code according to the information field screening strategy if the current intelligent contract code uploaded by the user side is detected; and
and the fuel limit prediction unit is used for inputting the current feature set into the automatic machine learning model for operation to obtain the Ether house fuel limit corresponding to the current intelligent contract code, and sending the Ether house fuel limit corresponding to the current intelligent contract code to a corresponding target receiving end.
8. The automated machine learning-based etherhouse fuel limit prediction device of claim 7, wherein the target network address acquisition unit comprises:
the first-level acquisition unit is used for acquiring all network addresses of all intelligent contracts issued on the ether house in a first-level webpage of the target website so as to form a first-level network address set;
traversing the next-level obtaining unit, which is used for accessing all second-level webpages adjacent to the first-level webpages and obtaining all network addresses of all intelligent contracts issued on the ether houses in the second-level webpages to form a second-level network address set; sequentially accessing all third-level webpages adjacent to the second-level webpages until all nth-level webpages adjacent to the (n-1) th-level webpages are accessed so as to respectively obtain a third-level network address set to an nth-level network address set; wherein the value of n is equal to the total webpage level number of the target website;
and the network address combination unit is used for forming network addresses of all intelligent contracts issued on the Etherns in the target website by the first-level network address set to the nth-level network address set.
9. A computer device comprising a memory, a processor and a computer program stored on the memory and executable on the processor, wherein the processor when executing the computer program implements the automatic machine learning-based etherhouse fuel limitation prediction method of any one of claims 1 to 6.
10. A computer-readable storage medium, characterized in that the computer-readable storage medium stores a computer program which, when executed by a processor, causes the processor to carry out the automatic machine learning-based etherhouse fuel limitation prediction method according to any one of claims 1 to 6.
Priority Applications (2)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010761121.3A CN111898766B (en) | 2020-07-31 | 2020-07-31 | Ether house fuel limitation prediction method and device based on automatic machine learning |
PCT/CN2020/118901 WO2021135467A1 (en) | 2020-07-31 | 2020-09-29 | Automated machine learning-based ethereum fuel restriction prediction method, apparatus, computer device, and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN202010761121.3A CN111898766B (en) | 2020-07-31 | 2020-07-31 | Ether house fuel limitation prediction method and device based on automatic machine learning |
Publications (2)
Publication Number | Publication Date |
---|---|
CN111898766A true CN111898766A (en) | 2020-11-06 |
CN111898766B CN111898766B (en) | 2023-02-24 |
Family
ID=73183474
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN202010761121.3A Active CN111898766B (en) | 2020-07-31 | 2020-07-31 | Ether house fuel limitation prediction method and device based on automatic machine learning |
Country Status (2)
Country | Link |
---|---|
CN (1) | CN111898766B (en) |
WO (1) | WO2021135467A1 (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112101553A (en) * | 2020-11-10 | 2020-12-18 | 鹏城实验室 | Network structure searching method and device, equipment and storage medium |
Families Citing this family (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN116611351B (en) * | 2023-07-19 | 2023-09-22 | 东方空间技术(山东)有限公司 | Rocket sublevel life assessment method, computer equipment and readable storage medium |
Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107967539A (en) * | 2017-11-29 | 2018-04-27 | 暨南大学 | The method for the fuel limitation merchandised on prediction ether mill based on machine learning and block chain technology |
CN108256052A (en) * | 2018-01-15 | 2018-07-06 | 成都初联创智软件有限公司 | Automobile industry potential customers' recognition methods based on tri-training |
CN110069690A (en) * | 2019-04-24 | 2019-07-30 | 成都市映潮科技股份有限公司 | A kind of theme network crawler method, apparatus and medium |
CN110751562A (en) * | 2019-10-29 | 2020-02-04 | 山东科技大学 | Training optimization method for foreigner time sequence prediction |
CN110942086A (en) * | 2019-10-30 | 2020-03-31 | 平安科技(深圳)有限公司 | Data prediction optimization method, device and equipment and readable storage medium |
CN111105160A (en) * | 2019-12-20 | 2020-05-05 | 北京工商大学 | Steel quality prediction method based on tendency heterogeneous bagging algorithm |
Family Cites Families (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109064325B (en) * | 2018-06-25 | 2020-07-24 | 浙江超脑时空科技有限公司 | Intelligent contract implementation method and device based on block chain |
CN109064656B (en) * | 2018-08-16 | 2020-09-11 | 黄竹磬 | Management system based on energy storage and power generation benefits of hydrogen fuel cell vehicle |
-
2020
- 2020-07-31 CN CN202010761121.3A patent/CN111898766B/en active Active
- 2020-09-29 WO PCT/CN2020/118901 patent/WO2021135467A1/en active Application Filing
Patent Citations (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN107967539A (en) * | 2017-11-29 | 2018-04-27 | 暨南大学 | The method for the fuel limitation merchandised on prediction ether mill based on machine learning and block chain technology |
CN108256052A (en) * | 2018-01-15 | 2018-07-06 | 成都初联创智软件有限公司 | Automobile industry potential customers' recognition methods based on tri-training |
CN110069690A (en) * | 2019-04-24 | 2019-07-30 | 成都市映潮科技股份有限公司 | A kind of theme network crawler method, apparatus and medium |
CN110751562A (en) * | 2019-10-29 | 2020-02-04 | 山东科技大学 | Training optimization method for foreigner time sequence prediction |
CN110942086A (en) * | 2019-10-30 | 2020-03-31 | 平安科技(深圳)有限公司 | Data prediction optimization method, device and equipment and readable storage medium |
CN111105160A (en) * | 2019-12-20 | 2020-05-05 | 北京工商大学 | Steel quality prediction method based on tendency heterogeneous bagging algorithm |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN112101553A (en) * | 2020-11-10 | 2020-12-18 | 鹏城实验室 | Network structure searching method and device, equipment and storage medium |
Also Published As
Publication number | Publication date |
---|---|
CN111898766B (en) | 2023-02-24 |
WO2021135467A1 (en) | 2021-07-08 |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN112101562B (en) | Implementation method and system of machine learning modeling process | |
JP7343568B2 (en) | Identifying and applying hyperparameters for machine learning | |
US10360517B2 (en) | Distributed hyperparameter tuning system for machine learning | |
CN107967539B (en) | Method for predicting fuel limit of transaction on ether house based on machine learning and block chain technology | |
JP7043596B2 (en) | Neural architecture search | |
US20180121535A1 (en) | Multiple record linkage algorithm selector | |
US20160012352A1 (en) | Data Processing Method and Computer System | |
KR101732319B1 (en) | IRIS: Goal-oriented big data business analytics framework | |
Cheng et al. | Efficient performance prediction for apache spark | |
CN111898766B (en) | Ether house fuel limitation prediction method and device based on automatic machine learning | |
CN114298323A (en) | Method and system for generating combined features of machine learning samples | |
CN110609870B (en) | Distributed data processing method and device, electronic equipment and storage medium | |
CA3137748A1 (en) | Method and apparatus for determining configuration knob of database | |
CN110706015A (en) | Advertisement click rate prediction oriented feature selection method | |
Bello Cruz et al. | Level bundle-like algorithms for convex optimization | |
CN114692978A (en) | Social media user behavior prediction method and system based on big data | |
WO2024139703A1 (en) | Object recognition model updating method and apparatus, electronic device, storage medium, and computer program product | |
Uc-Cetina et al. | Composition of web services using Markov decision processes and dynamic programming | |
CN112231481A (en) | Website classification method and device, computer equipment and storage medium | |
CN111859238A (en) | Method and device for predicting data change frequency based on model and computer equipment | |
Casimiro et al. | A probabilistic model checking approach to self-adapting machine learning systems | |
CN116383521A (en) | Subject word mining method and device, computer equipment and storage medium | |
CN110705889A (en) | Enterprise screening method, device, equipment and storage medium | |
US11922441B2 (en) | Method and system for event prediction using spatio-temporally sampled data | |
Sadman et al. | ADCR: An Adaptive TOOL to select” Appropriate Developer for Code Review” based on Code Context |
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 |