Method, apparatus and electronic device for optimizing machine learning model

Abstract

The application provides an optimization method and device of a machine learning model and an electronic device, relates to the technical field of cloud computing and intelligent operation and maintenance, and according to the embodiment of the application, when the machine learning model is changed, the configuration file and the training sample data of the machine learning model can be acquired, the machine learning model, the training sample features of the machine learning model and the running environment of the machine learning model are determined, the machine learning model can automatically train in the running environment, and the result related data is provided to a user, so as to provide a reference for optimization of the machine learning model. Artificial modification of various process related source codes is not needed according to the change of the machine learning model, the human and time costs of iteration of the machine learning model in the intelligent operation and maintenance scene are reduced, and the iteration efficiency is improved.

Description

Technical Field

[0001] This application relates to the fields of cloud computing and intelligent operation and maintenance technology, and in particular to a method, apparatus and electronic device for optimizing a machine learning model. Background Technology

[0002] Artificial intelligence and machine learning technologies are transforming entire industries, including the construction and operation of cloud services. Artificial Intelligence for IT Operations (AIOps) is a product of the combination of artificial intelligence and operations. By applying machine learning techniques, it performs real-time or offline analysis on massive amounts of operational data (including logs, business data, system data, etc.), enhancing the operational analysis capabilities of information technology (IT).

[0003] In the field of intelligent operations and maintenance (O&M), to improve the analytical accuracy and operational efficiency of machine learning models, iterative optimization of various machine learning models is necessary. Currently, in various solutions, changes to each machine learning model involved in the entire intelligent O&M scenario require modification of the entire process's related source code with the participation of developers to receive feedback. This significantly increases the development time and manpower costs for iterating machine learning models in intelligent O&M scenarios. Summary of the Invention

[0004] This application provides a method, apparatus, and electronic device for optimizing machine learning models, so as to reduce the human and time costs of iterating machine learning models in intelligent operation and maintenance scenarios and improve iteration efficiency.

[0005] In a first aspect, embodiments of this application provide a method for optimizing a machine learning model, the method comprising:

[0006] Obtain the configuration file and training sample data of the machine learning model; the configuration file includes the sample feature configuration file and the model configuration file of the machine learning model;

[0007] Based on the configuration file, the machine learning model, the features of the training samples for the machine learning model, and the operating environment of the machine learning model are determined.

[0008] In the operating environment, the machine learning model is trained using the features of the training samples to obtain the training results and related data of the machine learning model;

[0009] The results and related data are provided to users to serve as a reference for optimizing machine learning models.

[0010] Secondly, embodiments of this application provide an optimization apparatus for a machine learning model, the apparatus comprising:

[0011] The acquisition module is used to acquire the configuration file and training sample data of the machine learning model; the configuration file includes the sample feature configuration file and the model configuration file of the machine learning model;

[0012] The determination module is used to determine the machine learning model, the features of the machine learning model's training samples, and the machine learning model's operating environment based on the configuration file.

[0013] The training module is used to train a machine learning model using the features of training samples in the runtime environment, and to obtain data related to the training results of the machine learning model.

[0014] Provide modules to offer results-related data to users to provide a reference for optimizing machine learning models.

[0015] Thirdly, embodiments of this application provide an electronic device, including a memory, a processor, and a computer program stored in the memory, wherein the processor, when executing the computer program, implements the method described in any of the above-mentioned embodiments.

[0016] Fourthly, embodiments of this application provide a computer-readable storage medium storing a computer program, which, when executed by a processor, implements the method described in any of the above-mentioned embodiments.

[0017] Compared with the prior art, this application has the following advantages:

[0018] This application provides a method, apparatus, and electronic device for optimizing a machine learning model. When the machine learning model changes, it can obtain the model's configuration file and training sample data, determine the machine learning model, its training sample features, and its operating environment, enabling the model to automatically train within that environment. The resulting data is then provided to the user for optimization purposes. This eliminates the need for manual modification of source code for each process when the machine learning model changes, reducing the human and time costs of iterating machine learning models in intelligent operation and maintenance scenarios and improving iteration efficiency.

[0019] The above description is only an overview of the technical solution of this application. In order to better understand the technical means of this application, it can be implemented according to the contents of the specification. In order to make the above and other objects, features and advantages of this application more obvious and understandable, specific embodiments of this application are given below. Attached Figure Description

[0020] In the accompanying drawings, unless otherwise specified, the same reference numerals throughout the various drawings denote the same or similar parts or elements. These drawings are not necessarily drawn to scale. It should be understood that these drawings depict only some embodiments according to this application and should not be construed as limiting the scope of this application.

[0021] Figure 1 A schematic diagram illustrating the optimization of the machine learning model provided in this application;

[0022] Figure 2 This is a schematic diagram illustrating the generation of a configuration file according to an embodiment of this application;

[0023] Figure 3 This is a flowchart of an optimization method for a machine learning model according to an embodiment of this application;

[0024] Figure 4 This is a flowchart of an optimization method for a machine learning model according to an embodiment of this application;

[0025] Figure 5 This is a structural block diagram of an optimization apparatus for a machine learning model according to an embodiment of this application; and

[0026] Figure 6 This is a block diagram of an electronic device used to implement embodiments of this application. Detailed Implementation

[0027] In the following description, only certain exemplary embodiments are briefly described. As those skilled in the art will recognize, the described embodiments can be modified in various ways without departing from the concept or scope of this application. Therefore, the drawings and description are considered to be exemplary in nature and not restrictive.

[0028] To facilitate understanding of the technical solutions of the embodiments of this application, the relevant technologies of the embodiments of this application are described below. The following relevant technologies are optional solutions and can be combined with the technical solutions of the embodiments of this application in any way, and all of them fall within the protection scope of the embodiments of this application.

[0029] Figure 1This is a schematic diagram illustrating an application scenario for implementing the method of this application embodiment. This application embodiment can be applied to machine learning operations and maintenance scenarios within intelligent operations and maintenance scenarios. Machine learning operations and maintenance (MLOps) refers to the organic integration of machine learning system development and operation, automating each step of the machine learning model building process (including integration, testing, release, and deployment). Specifically, this embodiment can be implemented through a machine learning operations and maintenance system, which can be deployed on a separate server or on the same server as the machine learning model. The machine learning model involved can be used for fault detection and root cause analysis. Control flow represents control commands, and data flow represents data such as configuration files. The machine learning model can include support vector machine models, decision tree models, neural network models, etc. The machine learning operations and maintenance system in this embodiment includes: a configuration loader, a main control process, and a monitoring server. The configuration loader monitors and reads the configuration file of the machine learning model in real time, loads and records the configuration file, determines whether the configuration file has been updated based on its modification time, and triggers the corresponding interface of the main control process when the configuration file changes. The main control process reads multiple sets of hyperparameters of the machine learning model from the configuration file and utilizes the configuration file and multiple sets of hyperparameters (such as...) Figure 2 The “Configuration and Parameters” section shows how to configure machine learning models to obtain multiple machine learning models (such as...). Figure 2 The "running instance" shown in the diagram) and the training sample feature set are used to generate the running environment (e.g., the running environment corresponding to each running instance) based on the configuration file. Figure 2 The virtual environment or container shown is used to generate the corresponding virtual environment. Within this environment, multiple machine learning models are trained using the feature set of training samples. The machine learning models are then updated and iterated. The trained metrics, configuration information, and model hyperparameters are sent to the monitoring server. The monitoring server statistically analyzes the metrics and displays them on the user terminal. It also provides interfaces for exporting and deploying machine learning models, further improving downstream applications and optimization of these models. This enables lifecycle management of machine learning models in intelligent operation and maintenance scenarios.

[0030] The configuration file for the machine learning model is as follows: Figure 2As shown, the configuration files include the machine learning operations and maintenance system runtime configuration file MLconfig, the machine learning template configuration file MLtemplate, the sample feature configuration file MLfeature, and the model configuration file. The model configuration file includes the model parameter configuration file MLmodel and the model environment configuration file MLproject. These configuration files are related to the configuration, data version, feature engineering, model engineering, runtime environment (e.g., venv virtual environment, conda environment, or container Dockerfile), and program entry point of the machine learning operations and maintenance system. Configuration files can be in YAML or JSON format, etc. The machine learning operations and maintenance system runtime configuration file MLconfig controls various runtime parameter configurations for the entire system, including the current application task entry point, version check cycle, and parallelism, and also records the path information for MLtemplate, MLfeature, MLmodel, and MLproject. The template configuration file MLtemplate is used to configure the preprocessing rules for the source log data, including data version information. This data version information represents the version information of the training sample data. For data-driven technologies like machine learning, the accuracy of models built from different versions of data will vary. The sample feature configuration file MLfeature is related to feature engineering. Feature engineering refers to the process of transforming raw data into features that better express the essence of the problem, mainly including three sub-processes: feature extraction, feature construction, and feature selection. The model parameter configuration file MLmodel and the model environment configuration file MLproject are related to model engineering. Model engineering covers the entire lifecycle of machine learning models, from building, training and tuning, model deployment, evaluation, and use to providing feedback and monitoring on the results of using the machine learning models, ensuring standardization throughout the entire lifecycle of the machine learning model with minimal cost. In this embodiment, the machine learning operation and maintenance system only needs to be started once. Subsequent changes to data versions, feature engineering, model engineering, and the system software settings can all be implemented by modifying the configuration files.

[0031] In this embodiment, by separating data versioning, feature generation rules, and model building details from the implementation code and transmitting them to the machine learning operations and maintenance system via configuration files, the system can periodically detect configuration changes in key steps such as data versioning, feature engineering, and model engineering. It automatically iterates and optimizes the machine learning model, eliminating the need for manual configuration of each process for changes to the model. This significantly reduces labor costs and iteration costs in the intelligent operations and maintenance process, improving iteration efficiency. When applying the machine learning model to fault detection and root cause analysis, it can improve the performance optimization iteration efficiency of these applications.

[0032] This application provides a method for optimizing a machine learning model, such as... Figure 3 The flowchart shown is a method for optimizing a machine learning model according to an embodiment of this application, which may include:

[0033] Step S301: Obtain the configuration file and training sample data of the machine learning model; the configuration file includes the sample feature configuration file and the model configuration file of the machine learning model.

[0034] Step S302: Based on the configuration file, determine the machine learning model, the training sample features of the machine learning model, and the operating environment of the machine learning model.

[0035] Step S303: In the operating environment, train the machine learning model using the features of the training samples to obtain the training results related data of the machine learning model.

[0036] Step S304: Provide the relevant data of the results to the user for reference in optimizing the machine learning model.

[0037] The execution entity in this application embodiment can be a computing device, such as a server. The specific implementation method for obtaining the configuration file and training sample data of the machine learning model can be automatic acquisition according to a preset time period, or acquisition according to a received data acquisition instruction; this application embodiment does not limit this approach.

[0038] Among them, the sample feature configuration file is related to feature engineering. Feature engineering refers to the process of transforming raw data into features that better express the essence of the problem, mainly including three sub-processes: feature extraction, feature construction, and feature selection. The model configuration file includes the model parameter configuration file and the model environment configuration file, and is related to model engineering. Model engineering covers multiple steps of machine learning model building, from construction, training and tuning, model deployment, evaluation, and use to feedback and monitoring of the results. The configuration file format can be YAML or JSON, etc.

[0039] The configuration file is used to configure the data required for training the machine learning model, such as the machine learning model itself, the training sample set for training the machine learning model, and the runtime environment required for training the machine learning model. The training sample set is the collection of training sample features obtained by extracting features from the training sample data.

[0040] The runtime environment can be an isolated environment required for training machine learning models, such as a venv virtual environment, a conda environment, or a container Dockerfile.

[0041] Once the data required for training the machine learning model is configured, the runtime environment is loaded, the machine learning model is trained using the features of the training samples, and the training results are sent to the user terminal for display, providing a reference for further adjusting model parameters and optimizing the model.

[0042] This application provides a method for optimizing a machine learning model. When the machine learning model changes, the configuration file and training sample data of the machine learning model can be obtained to determine the machine learning model, its training sample features, and its operating environment. This allows the machine learning model to be automatically trained in the operating environment, and the resulting data is provided to the user for reference in optimizing the model. This eliminates the need for manual modification of the source code related to each process when the machine learning model changes, reducing the manpower and time costs of iterating machine learning models in intelligent operation and maintenance scenarios, and improving iteration efficiency.

[0043] The training sample features of the machine learning model in step S302 are obtained in the following way:

[0044] In one implementation, the training sample features of the machine learning model are determined based on a configuration file, including: obtaining sample feature generation rules from the sample feature configuration file, and determining the training sample features based on the sample feature generation rules and the training sample data.

[0045] Specifically, the sample feature generation rules are configured using the sample feature configuration file in the configuration file. These rules are then used to extract features from the training sample data, thus obtaining the training sample features. Depending on the functionality implemented by the machine learning model, different training sample data can be selected to obtain corresponding training sample features, which are then used as the data required for training the machine learning model.

[0046] In one implementation, the machine learning model and its runtime environment are determined based on a configuration file, including: obtaining hyperparameters and runtime environment-related information of the machine learning model from the model configuration file; determining the machine learning model based on the hyperparameters; and determining the runtime environment of the machine learning model based on the runtime environment-related information.

[0047] The model configuration file includes a parameter configuration file and a model environment configuration file. The model configuration file is related to the model project. The parameter configuration file is used to configure the hyperparameters of the machine learning model, which determines the specific machine learning model. The model environment configuration file is used to configure information related to the runtime environment of the machine learning model, including the type of runtime environment to build and the program entry function. The type of runtime environment can be determined according to specific needs; for example, depending on the needs of different machine learning models, appropriate virtual environments (venv), conda environments, or container Dockerfiles may be configured.

[0048] The following example further illustrates how to obtain hyperparameters and runtime environment information of a machine learning model from its configuration file:

[0049] In one implementation, the hyperparameters and runtime environment information of the machine learning model are obtained from the model configuration file. This includes: obtaining multiple sets of hyperparameters of the machine learning model and runtime environment information corresponding to each set of hyperparameters from the model configuration file; determining the machine learning model based on the hyperparameters; and determining the runtime environment of the machine learning model based on the runtime environment information. This includes: determining multiple machine learning models based on the multiple sets of hyperparameters; and determining the runtime environment corresponding to each of the multiple machine learning models based on the runtime environment information corresponding to each set of hyperparameters. The runtime environment is used to isolate the multiple machine learning models.

[0050] The model configuration file can configure multiple sets of hyperparameters for the machine learning model. Multiple machine learning models can be obtained based on these hyperparameters. The runtime environment information corresponding to each set of hyperparameters determines the runtime environment for each machine learning model. Updates and iterations are then performed within each machine learning model's respective runtime environment. Multiple machine learning models can be trained in parallel within their respective runtime environments, improving training efficiency. The runtime environment can be a virtual environment or a container. By applying virtual environment and container technologies, an independent runtime environment is constructed for each machine learning model, achieving isolation between models and ensuring good stability for the entire system. The resulting machine learning models are also reproducible and portable, suitable for various production environments in intelligent operation and maintenance scenarios.

[0051] In one implementation, obtaining the configuration file and training sample data of the machine learning model includes: obtaining the configuration file and source log data according to a preset time period, and determining that the configuration file and source log data have been updated; preprocessing the source log data to obtain the training sample data.

[0052] In practical applications, the configuration file and training sample data of the machine learning model are periodically retrieved according to a preset time period. Based on the modification time of the configuration file and training sample data, it is determined whether the configuration file and source log data have been updated compared to before. If there are updates, the source log data is preprocessed to obtain the training sample data. Preprocessing may include cleaning, extracting, and matching the source log data according to preprocessing rules. The preset time period can be set according to specific needs, for example, 2 seconds or 5 seconds.

[0053] In one implementation, the configuration file also includes a template configuration file; the method further includes: obtaining data version information of the training sample data from the template configuration file; and the result-related data includes the data version information.

[0054] In practical applications, template configuration files are used to configure the preprocessing rules for source log data. This includes data version information, which represents the version of the training sample data. For machine learning models, the accuracy of models built based on different versions of data varies. When displaying the model's training results, the data version information can be provided to the user as part of the results-related data, allowing the user to determine the model's optimization strategy based on this information.

[0055] In one implementation, providing result-related data to the user includes: sending the result-related data to a monitoring server, and the monitoring server displaying the result-related data through the user's terminal; the result-related data includes the hyperparameters of the machine learning model and training result metrics.

[0056] In practical applications, the training process and results of the model can be monitored by a monitoring server. Result-related data is sent to the monitoring server, which then sends the data to the user terminal based on the query command, where it is displayed on the user's terminal. This result-related data can include training results related to the machine learning model. For example, if the machine learning model is a classification model, the training results data could include classification accuracy. If the upper-layer application of this classification model is log denoising in an intelligent operations and maintenance scenario, the training results data could also include the denoising rate. The specific functions implemented by the machine learning model will result in different result-related data. Additionally, the result-related data also includes input information such as the machine learning model's hyperparameters, data version information, and configuration file version.

[0057] To more clearly illustrate the technical concept of this application, a detailed description is provided below through specific embodiments. Figure 4 This is a flowchart illustrating a method for optimizing a machine learning model according to an embodiment of this application. The method includes:

[0058] Step S401: The main control process checks the last modification time of all configuration files and source log data at regular intervals.

[0059] Step S402: Determine whether the configuration file and source log data have changed. If yes, proceed to steps S403-S407; otherwise, return to step S401.

[0060] In step S403, the master control process preprocesses the source log data, records the last modification time of the data, and reads the data version information from the MLtemplate file.

[0061] Step S404: Read and load the feature extraction and generation rules from the MLfeature file; read and load the hyperparameters of the machine learning model from the MLmodel file.

[0062] Step S405: Read and load the runtime environment information and program entry point from the MLproject file, and generate the corresponding virtual environment or container.

[0063] In step S406, the main control process trains and validates the corresponding machine learning model in a virtual environment or container based on the current configuration information, and sends the parameters, metrics, and analysis results to the monitoring server. Because the virtual environment or container isolates the running environment of each machine learning model, multiple machine learning models can be trained in parallel for different configurations and hyperparameter selections.

[0064] In step S407, the monitoring server displays parameters, indicators, and analysis results through the user terminal, and provides an interface for deploying and managing machine learning models; the main control process returns to step S401 and re-executes the entire process.

[0065] Corresponding to the application scenarios and methods provided in the embodiments of this application, the embodiments of this application also provide an optimization apparatus for machine learning models. For example... Figure 5 The diagram shown is a structural block diagram of a machine learning model optimization device according to an embodiment of this application, which may include:

[0066] The acquisition module 501 is used to acquire the configuration file and training sample data of the machine learning model; the configuration file includes the sample feature configuration file and the model configuration file of the machine learning model;

[0067] Module 502 is used to determine the machine learning model, the features of the training samples of the machine learning model, and the running environment of the machine learning model based on the configuration file.

[0068] Training module 503 is used to train a machine learning model using the features of training samples in the runtime environment and obtain training result data related to the machine learning model.

[0069] Module 504 is provided to provide users with relevant data on the results, which can be used as a reference for optimizing machine learning models.

[0070] This application provides an optimization device for a machine learning model. When the machine learning model changes, it can obtain the configuration file and training sample data of the machine learning model, determine the machine learning model, the features of the training samples, and the operating environment of the machine learning model. This allows the machine learning model to be automatically trained in the operating environment, and the resulting data is provided to the user for reference in optimizing the machine learning model. It eliminates the need for manual modification of the source code related to each process when the machine learning model changes, reducing the manpower and time costs of iterating machine learning models in intelligent operation and maintenance scenarios, and improving iteration efficiency.

[0071] In one implementation, when determining the training sample features of the machine learning model based on the configuration file, the determining module 502 is used to: obtain the sample feature generation rules from the sample feature configuration file, and determine the training sample features based on the sample feature generation rules and the training sample data.

[0072] In one implementation, when determining the machine learning model and its runtime environment based on the configuration file, the determining module 502 is used to: obtain the hyperparameters and runtime environment-related information of the machine learning model from the model configuration file; determine the machine learning model based on the hyperparameters; and determine the runtime environment of the machine learning model based on the runtime environment-related information.

[0073] In one implementation, when the acquisition module 502 obtains hyperparameters and runtime environment information of the machine learning model from the model configuration file, it is used to: obtain multiple sets of hyperparameters of the machine learning model and runtime environment information corresponding to each set of hyperparameters from the model configuration file; when the determination module 502 determines the machine learning model based on the hyperparameters and determines the runtime environment of the machine learning model based on the runtime environment information, it is used to: determine multiple machine learning models based on the multiple sets of hyperparameters and determine the runtime environment corresponding to each of the multiple machine learning models based on the runtime environment information corresponding to each set of hyperparameters, wherein the runtime environment is used to isolate the multiple machine learning models.

[0074] In one implementation, the acquisition module 501 is used to: acquire configuration files and source log data according to a preset time period, determine that the configuration files and source log data have been updated, and preprocess the source log data to obtain training sample data.

[0075] In one implementation, the configuration file further includes a template configuration file; the device is also used to: obtain data version information of the training sample data from the template configuration file; the result-related data includes the data version information.

[0076] In one implementation, module 504 is provided for: sending result-related data to a monitoring server, and the monitoring server displaying the result-related data through a user terminal; the result-related data includes hyperparameters of the machine learning model and training result metrics.

[0077] The functions of each module in each device in the embodiments of this application can be found in the corresponding description in the above method, and they have corresponding beneficial effects, which will not be repeated here.

[0078] Figure 6 This is a block diagram of an electronic device used to implement embodiments of this application. For example... Figure 6 As shown, the electronic device includes a memory 610 and a processor 620. The memory 610 stores a computer program that can run on the processor 620. When the processor 620 executes the computer program, it implements the methods described in the above embodiments. The number of memories 610 and processors 620 can be one or more.

[0079] The electronic device also includes:

[0080] The communication interface 630 is used to communicate with external devices and perform data exchange and transmission.

[0081] If the memory 610, processor 620, and communication interface 630 are implemented independently, they can be interconnected via a bus to communicate with each other. This bus can be an Industry Standard Architecture (ISA) bus, a Peripheral Component Interconnect (PCI) bus, or an Extended Industry Standard Architecture (EISA) bus, etc. This bus can be divided into address bus, data bus, control bus, etc. For ease of representation, Figure 6 The bus is represented by a single thick line, but this does not mean that there is only one bus or one type of bus.

[0082] Optionally, in a specific implementation, if the memory 610, processor 620, and communication interface 630 are integrated on a single chip, then the memory 610, processor 620, and communication interface 630 can communicate with each other through an internal interface.

[0083] This application provides a computer-readable storage medium storing a computer program that, when executed by a processor, implements the method provided in this application.

[0084] This application also provides a chip including a processor for calling and executing instructions stored in a memory, causing a communication device with the chip installed to perform the method provided in this application.

[0085] This application also provides a chip, including: an input interface, an output interface, a processor, and a memory. The input interface, output interface, processor, and memory are connected through an internal connection path. The processor is used to execute code in the memory. When the code is executed, the processor is used to execute the method provided in the application embodiment.

[0086] It should be understood that the aforementioned processor can be a Central Processing Unit (CPU), or 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, etc. General-purpose processors can be microprocessors or any conventional processor. It is worth noting that the processor can be a processor supporting Advanced Reduced Instruction Set Machines (ARM) architecture.

[0087] Further, optionally, the aforementioned memory may include read-only memory and random access memory. The memory may be volatile memory or non-volatile memory, or may include both. Non-volatile memory may include read-only memory (ROM), programmable read-only memory (PROM), erasable programmable read-only memory (EPROM), electrically erasable programmable read-only memory (EEPROM), or flash memory. Volatile memory may include random access memory (RAM), which serves as an external cache. By way of example, but not limitation, many forms of RAM are available. Examples include Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), Synchronous DRAM (SDRAM), Double Data Rate SDRAM (DDR SDRAM), Enhanced Synchronous DRAM (ESDRAM), Synchronous Link DRAM (SLDRAM), and Direct Rambus RAM (DR RAM).

[0088] In the above embodiments, implementation can be achieved, in whole or in part, through software, hardware, firmware, or any combination thereof. When implemented in software, it can be implemented, in whole or in part, as a computer program product. A computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, all or part of the processes or functions according to this application are generated. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transferred from one computer-readable storage medium to another.

[0089] In the description of this specification, the references to terms such as "one embodiment," "some embodiments," "example," "specific example," or "some examples," etc., indicate that a specific feature, structure, material, or characteristic described in connection with that embodiment or example is included in at least one embodiment or example of this application. Furthermore, the specific features, structures, materials, or characteristics described may be combined in any suitable manner in one or more embodiments or examples. Moreover, without contradiction, those skilled in the art can combine and integrate the different embodiments or examples described in this specification, as well as the features of those different embodiments or examples.

[0090] Furthermore, the terms "first" and "second" are used for descriptive purposes only and should not be construed as indicating or implying relative importance or implicitly specifying the number of technical features indicated. Thus, a feature defined as "first" or "second" may explicitly or implicitly include at least one of that feature. In the description of this application, "a plurality of" means two or more, unless otherwise explicitly specified.

[0091] Any process or method described in the flowchart or otherwise herein can be understood as representing a module, segment, or portion of code comprising one or more executable instructions for implementing a particular logical function or process. Furthermore, the scope of the preferred embodiments of this application includes additional implementations in which functions may be performed not in the order shown or discussed, including substantially simultaneously or in reverse order depending on the functionality involved.

[0092] The logic and / or steps described in the flowchart or otherwise herein, for example, can be considered as a sequenced list of executable instructions for implementing logical functions, and can be embodied in any computer-readable medium for use by, or in conjunction with, an instruction execution system, apparatus or device (such as a computer-based system, a processor-included system or other system that can fetch and execute instructions from, an instruction execution system, apparatus or device).

[0093] It should be understood that various parts of this application can be implemented using hardware, software, firmware, or a combination thereof. In the above embodiments, multiple steps or methods can be implemented using software or firmware stored in memory and executed by a suitable instruction execution system. All or part of the steps of the methods in the above embodiments can be implemented by a program instructing related hardware, the program being stored in a computer-readable storage medium, which, when executed, includes one or a combination of the steps of the method embodiments.

[0094] Furthermore, the functional units in the various embodiments of this application can be integrated into a processing module, or each unit can exist physically separately, or two or more units can be integrated into a module. The integrated module can be implemented in hardware or as a software functional module. If the integrated module is implemented as a software functional module and sold or used as an independent product, it can also be stored in a computer-readable storage medium. This storage medium can be a read-only memory, a disk, or an optical disk, etc.

[0095] The above description is merely an exemplary embodiment of this application, but the scope of protection of this application is not limited thereto. Any person skilled in the art can easily conceive of various variations or substitutions within the technical scope described in this application, and these should all be included within the scope of protection of this application. Therefore, the scope of protection of this application should be determined by the scope of the claims.

Claims

1. A method for optimizing a machine learning model, the method comprising: The method includes: Obtain the configuration file and training sample data of the machine learning model; the configuration file includes the sample feature configuration file and the model configuration file of the machine learning model; Based on the configuration file, the machine learning model, the training sample features of the machine learning model, and the operating environment of the machine learning model are determined. In the operating environment, the machine learning model is trained using the features of the training samples to obtain training result-related data of the machine learning model; The results and related data will be provided to the user to serve as a reference for optimizing the machine learning model; The process of determining the machine learning model and its operating environment based on the configuration file includes: obtaining multiple sets of hyperparameters of the machine learning model and corresponding operating environment information from the model configuration file; determining multiple machine learning models based on the multiple sets of hyperparameters; and determining the operating environments corresponding to the multiple machine learning models based on the corresponding operating environment information, wherein the operating environments are used to isolate the multiple machine learning models.

2. The method of claim 1, wherein, Based on the configuration file, the training sample features of the machine learning model are determined, including: The sample feature generation rules are obtained from the sample feature configuration file, and the training sample features are determined based on the sample feature generation rules and the training sample data.

3. The method of claim 1, wherein, The process of obtaining the configuration file and training sample data for the machine learning model includes: The configuration file and source log data are obtained according to a preset time period to determine that the configuration file and source log data have been updated; The source log data is preprocessed to obtain the training sample data.

4. The method according to claim 1, characterized in that, The configuration file also includes a template configuration file; the method further includes: The data version information of the training sample data is obtained from the template configuration file; the result-related data includes the data version information.

5. The method of claim 1, wherein, Providing the relevant data of the results to the user includes: The result-related data is sent to the monitoring server, which displays the result-related data through the user terminal; the result-related data includes the hyperparameters and training result metrics of the machine learning model.

6. An apparatus for optimizing a machine learning model, the apparatus comprising: The device includes: The acquisition module is used to acquire the configuration file and training sample data of the machine learning model; the configuration file includes the sample feature configuration file and the model configuration file of the machine learning model; The determination module is used to determine the machine learning model, the training sample features of the machine learning model, and the operating environment of the machine learning model based on the configuration file. The training module is used to train the machine learning model using the features of the training samples in the operating environment, and obtain training result-related data of the machine learning model. A module is provided to provide the user with the relevant data of the results, so as to provide a reference for the optimization of the machine learning model; The determining module is used to obtain multiple sets of hyperparameters of the machine learning model and the runtime environment information corresponding to the multiple sets of hyperparameters from the model configuration file; determine multiple machine learning models based on the multiple sets of hyperparameters; and determine the runtime environment corresponding to the multiple machine learning models based on the runtime environment information corresponding to the multiple sets of hyperparameters, wherein the runtime environment is used to isolate the multiple machine learning models.

7. An electronic device, characterized in that, It includes a memory, a processor, and a computer program stored in the memory, wherein the processor, when executing the computer program, implements the method of any one of claims 1-5.

8. A computer-readable storage medium storing a computer program that, when executed by a processor, implements the method of any one of claims 1-5.

Images