CN113657481A - Model construction system and method - Google Patents
Model construction system and method Download PDFInfo
- Publication number
- CN113657481A CN113657481A CN202110931700.2A CN202110931700A CN113657481A CN 113657481 A CN113657481 A CN 113657481A CN 202110931700 A CN202110931700 A CN 202110931700A CN 113657481 A CN113657481 A CN 113657481A
- Authority
- CN
- China
- Prior art keywords
- data
- module
- target
- model
- sample data
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 52
- 238000010276 construction Methods 0.000 title claims abstract description 11
- 238000012545 processing Methods 0.000 claims abstract description 29
- 238000012549 training Methods 0.000 claims abstract description 22
- 238000000605 extraction Methods 0.000 claims abstract description 15
- 238000000926 separation method Methods 0.000 claims abstract description 14
- 238000001514 detection method Methods 0.000 claims abstract description 10
- 230000008569 process Effects 0.000 abstract description 24
- 230000008859 change Effects 0.000 description 4
- 230000001419 dependent effect Effects 0.000 description 4
- 230000000694 effects Effects 0.000 description 4
- 238000012216 screening Methods 0.000 description 4
- 230000009471 action Effects 0.000 description 3
- 238000006243 chemical reaction Methods 0.000 description 2
- 238000011161 development Methods 0.000 description 2
- 238000005516 engineering process Methods 0.000 description 2
- 230000001788 irregular Effects 0.000 description 2
- 230000002159 abnormal effect Effects 0.000 description 1
- 238000013473 artificial intelligence Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000013480 data collection Methods 0.000 description 1
- 238000013075 data extraction Methods 0.000 description 1
- 238000003066 decision tree Methods 0.000 description 1
- 230000003247 decreasing effect Effects 0.000 description 1
- 238000009795 derivation Methods 0.000 description 1
- 238000010586 diagram Methods 0.000 description 1
- 238000010801 machine learning Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000002265 prevention Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
Images
Classifications
-
- 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
-
- 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
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- 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
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)
- Computer Vision & Pattern Recognition (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Bioinformatics & Computational Biology (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Artificial Intelligence (AREA)
- Life Sciences & Earth Sciences (AREA)
- Business, Economics & Management (AREA)
- Accounting & Taxation (AREA)
- Development Economics (AREA)
- Economics (AREA)
- Finance (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 application provides a model construction system and a model construction method, and belongs to the technical field of model construction. The system comprises: the data acquisition module is used for acquiring sample data of at least one data source; the data detection module is used for detecting sample data of at least one data source based on a preset quality standard corresponding to the data source and taking the sample data meeting the preset quality standard as target sample data; the characteristic extraction module is used for extracting the data characteristics of the target sample data; the characteristic analysis module is used for selecting the data characteristics which accord with preset characteristic indexes from the data characteristics as target characteristics; the characteristic processing module is used for determining the distribution characteristics of the target characteristics and performing box separation processing on the target characteristics based on the distribution characteristics to obtain a plurality of characteristic boxes; and the model training module is used for training the target model based on the plurality of feature boxes. And standardization and streamlining of the model construction process are realized.
Description
Technical Field
The present application relates to the field of model building technologies, and in particular, to a model building system and method.
Background
With the development of the technology in the fields of artificial intelligence and machine learning, data modeling provides a new technical means for data enabling. By modeling and analyzing the mass data, business support and high-accuracy strategy recommendation can be provided for multiple scenes such as accurate marketing, risk prevention and control, financial credit and the like.
At present, in the wind control modeling process of the internet financial industry, links such as data acquisition, data processing, model training and the like are generally performed separately, and each link is operated by a developer in charge of the link on the machine.
However, the modeling by the method is insufficient in process and standardization, and the problems of weak model effect and low accuracy caused by missing a certain link, irregular operation and high subjectivity of a developer exist.
Disclosure of Invention
The embodiment of the application aims to provide a model construction system and a model construction method so as to solve the problems of weak model effect and low accuracy caused by insufficient flow and standardization of the current modeling process. The specific technical scheme is as follows:
in a first aspect, a model building system is provided, the system comprising: the system comprises a data acquisition module, a data detection module, a feature extraction module, a feature analysis module, a feature processing module and a model training module;
the data acquisition module is used for acquiring sample data of at least one data source;
the data detection module is used for detecting the sample data of at least one data source based on a preset quality standard corresponding to the data source and taking the sample data meeting the preset quality standard as target sample data;
the characteristic extraction module is used for extracting the data characteristics of the target sample data;
the characteristic analysis module is used for selecting the data characteristics which accord with preset characteristic indexes from the data characteristics as target characteristics;
the characteristic processing module is used for determining the distribution characteristics of the target characteristics and performing box separation processing on the target characteristics based on the distribution characteristics to obtain a plurality of characteristic boxes;
and the model training module is used for training a target model based on a plurality of feature boxes.
In one possible embodiment, the system further comprises: a model analysis module;
and the model analysis module is used for analyzing the execution condition of each module based on the analysis index corresponding to the module so as to obtain the analysis result corresponding to the module.
In one possible embodiment, the system further comprises: a report generation module;
the report generation module is used for generating an analysis report based on the analysis result and a preset report format.
In one possible embodiment, the system further comprises: and the model deployment module is used for determining a configuration file of the target model and performing online deployment on the target model based on the configuration file.
In one possible implementation, the feature extraction module includes: extracting submodules and deriving the submodules;
the extraction submodule is used for extracting basic features of the target sample data and taking the basic features as the data features;
and the derivative submodule is used for processing the basic characteristics to obtain derivative characteristics, and taking the derivative characteristics as the data characteristics.
In a second aspect, a model building method is provided, the method comprising:
collecting sample data of at least one data source;
detecting sample data of at least one data source based on a preset quality standard corresponding to the data source, and taking the sample data meeting the preset quality standard as target sample data;
extracting data characteristics of the target sample data, and selecting the data characteristics which accord with preset characteristic indexes from the data characteristics as target characteristics;
determining the distribution characteristics of the target characteristics, and performing box separation processing on the target characteristics based on the distribution characteristics to obtain a plurality of characteristic boxes;
training a target model based on a plurality of the feature boxes until the target model converges.
In one possible embodiment, the method further comprises:
and aiming at each module, analyzing the execution condition of the module based on the analysis index corresponding to the module to obtain the analysis result corresponding to the module.
In one possible embodiment, the method further comprises:
and generating an analysis report based on the analysis result and a preset report format.
In one possible embodiment, the method further comprises:
determining a configuration file of the target model;
and performing online deployment on the target model based on the configuration file.
In one possible embodiment, the extracting the data feature of the target sample data includes:
extracting basic features of the target sample data, and taking the basic features as the data features;
and processing the basic features to obtain derivative features, and taking the derivative features as the data features.
The embodiment of the application has the following beneficial effects:
the embodiment of the application provides a model building system and a method, wherein the system comprises: the data acquisition module is used for acquiring sample data of at least one data source; the data detection module is used for detecting the sample data of at least one data source based on a preset quality standard corresponding to the data source and taking the sample data meeting the preset quality standard as target sample data; the characteristic extraction module is used for extracting the data characteristics of the target sample data; the characteristic analysis module is used for selecting the data characteristics which accord with preset characteristic indexes from the data characteristics as target characteristics; the characteristic processing module is used for determining the distribution characteristics of the target characteristics and performing box separation processing on the target characteristics based on the distribution characteristics to obtain a plurality of characteristic boxes; and the model training module is used for training the target model based on the plurality of feature boxes.
Namely, the modules required in model building are integrated in one system, so that the model building process is standardized and streamlined, and the influence on the performance of the finally built model due to missing of a certain link caused by manual operation in the model building process is avoided.
Of course, not all advantages described above need to be achieved at the same time in the practice of any one product or method of the present application.
Drawings
In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly described below, and it is obvious for those skilled in the art to obtain other drawings without inventive exercise.
Fig. 1 is a schematic structural diagram of a model building system according to an embodiment of the present application;
fig. 2 is a flowchart of a model building method according to an embodiment of the present application.
Detailed Description
The technical solutions in the embodiments of the present application will be clearly and completely described below with reference to the drawings in the embodiments of the present application, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. 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 application.
At present, in the wind control modeling process of the internet financial industry, links such as data acquisition, data processing, model training and the like are generally performed separately, and each link is operated by a developer in charge of the link. However, the modeling by the method is insufficient in process and standardization, and the problems of weak model effect and low accuracy caused by missing a certain link, irregular operation and high subjectivity of a developer exist. Therefore, the embodiment of the application provides a model construction method.
As shown in fig. 1, a model building system provided for an embodiment of the present application includes: the system comprises a data acquisition module, a data detection module, a feature extraction module, a feature analysis module, a feature processing module and a model training module;
the data acquisition module is used for acquiring sample data of at least one data source;
the data detection module is used for detecting the sample data of at least one data source based on a preset quality standard corresponding to the data source and taking the sample data meeting the preset quality standard as target sample data;
the characteristic extraction module is used for extracting the data characteristics of the target sample data;
the characteristic analysis module is used for selecting the data characteristics which accord with preset characteristic indexes from the data characteristics as target characteristics;
the characteristic processing module is used for determining the distribution characteristics of the target characteristics and performing box separation processing on the target characteristics based on the distribution characteristics to obtain a plurality of characteristic boxes;
and the model training module is used for training a target model based on a plurality of feature boxes.
In yet another embodiment of the present application, the system further comprises: a model analysis module;
and the model analysis module is used for analyzing the execution condition of each module based on the analysis index corresponding to the module so as to obtain the analysis result corresponding to the module.
In yet another embodiment of the present application, the system further comprises: a report generation module;
the report generation module is used for generating an analysis report based on the analysis result and a preset report format.
In yet another embodiment of the present application, the system further comprises: and the model deployment module is used for determining a configuration file of the target model and performing online deployment on the target model based on the configuration file.
In another embodiment of the present application, the feature extraction module includes: extracting submodules and deriving the submodules;
the extraction submodule is used for extracting basic features of the target sample data and taking the basic features as the data features;
and the derivative submodule is used for processing the basic characteristics to obtain derivative characteristics, and taking the derivative characteristics as the data characteristics.
In an embodiment of the present application, the system includes: the data acquisition module is used for acquiring sample data of at least one data source; the data detection module is used for detecting the sample data of at least one data source based on a preset quality standard corresponding to the data source and taking the sample data meeting the preset quality standard as target sample data; the characteristic extraction module is used for extracting the data characteristics of the target sample data; the characteristic analysis module is used for selecting the data characteristics which accord with preset characteristic indexes from the data characteristics as target characteristics; the characteristic processing module is used for determining the distribution characteristics of the target characteristics and performing box separation processing on the target characteristics based on the distribution characteristics to obtain a plurality of characteristic boxes; and the model training module is used for training the target model based on the plurality of feature boxes. Namely, the modules required in model building are integrated in one system, so that the model building process is standardized and streamlined, and the influence on the performance of the finally built model due to missing of a certain link caused by manual operation in the model building process is avoided.
Optionally, an embodiment of the present application further provides a model construction method, as shown in fig. 2, which includes the following specific steps.
S101, collecting sample data of at least one data source.
The embodiment of the application is applied to a wind control modeling scene in the internet financial industry, and the sample data is generally attribute data (such as user identification, basic information data, common loan behavior data and the like) of a historical user. In practical applications, due to differences in business or product lines, there are generally a plurality of data sources for collecting sample data. In the data collection process, the required data can be extracted as sample data by setting a screening label (such as a time range, a user channel or a new customer/old customer) so as to improve the data extraction efficiency.
S102, aiming at the sample data of at least one data source, detecting the sample data based on a preset quality standard corresponding to the data source, and taking the sample data meeting the preset quality standard as target sample data.
In the embodiment of the application, corresponding quality standards are set in advance according to different data sources, and the quality standards are used for controlling the quality of sample data. And detecting sample data corresponding to the data source based on preset quality standards corresponding to different data sources, and taking the sample data meeting the preset quality standards as target sample data, so that the data quality is improved, and the modeling reliability is ensured. For example, for repayment data, data with a loss rate less than a preset threshold is determined as meeting a preset quality standard. The quality standard can also be set according to the annual/monthly change trend of the overdue rate, the annual/monthly change trend of the variable loss rate, the abnormal variable detection result and the like of the user.
S103, extracting the data characteristics of the target sample data, and selecting the data characteristics which accord with preset characteristic indexes from the data characteristics as target characteristics.
In the embodiment of the application, the data feature includes a basic feature and a derivative feature, and the preset feature index may be a prediction capability of the data feature, stability of the data feature, and the like. And after the data characteristics of the target sample data are extracted, detecting the data characteristics, and taking the data characteristics which accord with preset characteristic indexes as target characteristics. And screening out data characteristics with weak prediction capability and low stability through characteristic indexes.
In an implementation manner of the embodiment of the present application, the data features of the target sample data may be extracted through the following steps:
step one, extracting basic features of the target sample data, and taking the basic features as the data features;
and step two, processing the basic features to obtain derivative features, and taking the derivative features as the data features.
In this embodiment, the basic features refer to features that can be directly extracted, and first extract basic features of target sample data, and then process the basic features based on a preset rule to obtain derived features, for example: performing summation derivation on basic characteristics with similar business meanings; or the ratio of the recent characteristics to the future characteristics to generate derivative characteristics of the reaction rate of change; or comparing whether the values of the two basic features are consistent or not to generate a derivative feature reflecting the matching degree, and finally, taking the basic feature and the derivative feature as data features. By the scheme, the basic characteristics and the derived characteristics can be used as data characteristics, so that the data characteristics can more comprehensively reflect the situation of sample data.
S104, determining the distribution characteristics of the target characteristics, and performing box separation processing on the target characteristics based on the distribution characteristics to obtain a plurality of characteristic boxes.
In the embodiment of the present application, the binning processing refers to grouping the target features, for example, the age range of the platform user is 20-50 years, and the grouping may be set as: [19.5, 30.5], [30.5, 40.5], [40.5, 50.5], and then, WOE (Evidence Weight) values corresponding to the bins are calculated, the WOE values being characteristic transformation modes capable of reflecting the relationship between independent variables and dependent variables of the model.
The binning mode in the embodiment of the application comprises one or more of decision tree binning, chi-square binning, coarse binning iteration and other binning algorithms, and the appropriate binning mode is selected according to different data distribution characteristics. For example, for continuous data, firstly carrying out coarse binning on the data, then merging the data according to the prediction capability of the binned data, and repeating iteration to obtain an optimal binning result; for the category type data, when the categories are more, the grouping can be combined according to overdue performance of the category users, and then the card square is selected for box division; for data with less values, each data can be used as a box. By the box separation of the scheme, the high prediction capability of the characteristics can be ensured, and meanwhile, the monotonicity of the WOE variable can be considered, namely after the box separation is carried out according to the box separation mode, the variable is subjected to WOE conversion, and the variable and the dependent variable are in a monotonically decreasing relation.
And S105, training a target model based on the plurality of feature boxes until the target model converges.
In the embodiment of the application, after the plurality of feature boxes are obtained, the target model is trained by using the plurality of feature boxes until the target model converges.
In the embodiment of the application, sample data of at least one data source is collected; then, detecting sample data of at least one data source based on a preset quality standard corresponding to the data source, and taking the sample data meeting the preset quality standard as target sample data; extracting data characteristics of the target sample data, and selecting the data characteristics which accord with preset characteristic indexes from the data characteristics as target characteristics; determining the distribution characteristics of the target characteristics, and performing box separation processing on the target characteristics based on the distribution characteristics to obtain a plurality of characteristic boxes; training a target model based on a plurality of the feature boxes until the target model converges. According to the method and the device, the model building process is standardized and streamlined, and the phenomenon that certain link is missed due to manual operation in the model building process to influence the performance of the finally built model is avoided.
In yet another embodiment of the present application, the method further comprises the steps of:
and aiming at each module, analyzing the execution condition of the module based on the analysis index corresponding to the module to obtain the analysis result corresponding to the module.
In this embodiment of the present application, an analysis index corresponding to each module is preset, and after the execution of each module is completed, the execution condition of the module is analyzed according to the analysis index corresponding to the module, so as to obtain an analysis result, where the analysis result may include but is not limited to: model overview, in-mold variable analysis, variable correlation, variable mean, in-mold variable distribution, prediction score distribution and variable trend analysis. According to the scheme, the analysis result can be automatically output after each module is completely executed, and manual analysis is not needed, so that developers can conveniently know problems existing in the model building process.
In yet another embodiment of the present application, the method further comprises the steps of:
and generating an analysis report based on the analysis result and a preset report format.
In the embodiment of the application, a report format may be preset, and the analysis result is typeset according to the preset report format to generate the analysis report. The analysis report may include a plurality of sub-reports, each sub-report includes a category of content, and the content of the analysis report specifically includes: an index of the model for understanding basic properties of the model (including subject contents of the model, including model name, date, development path, data set division mode, independent variable, dependent variable, main model parameter value, and the like); model overview (an index showing the overall performance of the model, such as accuracy and stability of the model); a variable screening process (showing the process of screening variables at each step); analyzing the model entering variable (information such as effectiveness, contribution degree and significance of the model entering variable); a model formula; variable correlation (for evaluating correlation between variables); a variable mean value; distribution of in-mold variables; predicting a score distribution; and (4) variable trend analysis (used for displaying the change trend of the dependent variable along with the independent variable).
The tuning direction can be determined by analyzing the contents in the report, for example, by analyzing the contents of the contribution degrees of the variables, if it is found that there is a high correlation between the variables x1 and x2, and the contribution degree of x2 to the model is small, the tuning direction can be obtained by cross-analyzing the sub-report where x1 is located and the sub-report where x2 is located: the x2 variable is removed. Therefore, the problem of multiple collinearity of the model caused by high correlation is avoided, and the effect of the model is favorably improved.
By the aid of the method and the device, all analysis data in the model building process can be displayed in a centralized mode in one analysis report, and developers can conveniently know all conditions in the model building process.
In yet another embodiment of the present application, the method further comprises the steps of:
step one, determining a configuration file of the target model;
and secondly, performing online deployment on the target model based on the configuration file.
In the embodiment of the application, after the target model is trained, the configuration file of the target model can be determined, and then the target model is deployed online based on the configuration file, so that the model is automatically online, and the online efficiency of the model is improved.
In the embodiment of the application, sample data of at least one data source is collected; then, detecting sample data of at least one data source based on a preset quality standard corresponding to the data source, and taking the sample data meeting the preset quality standard as target sample data; extracting data characteristics of the target sample data, and selecting the data characteristics which accord with preset characteristic indexes from the data characteristics as target characteristics; determining the distribution characteristics of the target characteristics, and performing box separation processing on the target characteristics based on the distribution characteristics to obtain a plurality of characteristic boxes; training a target model based on a plurality of the feature boxes until the target model converges. According to the method and the device, the model building process is standardized and streamlined, and the phenomenon that certain link is missed due to manual operation in the model building process to influence the performance of the finally built model is avoided.
It is noted that, in this document, relational terms such as "first" and "second," and the like, may be used solely to distinguish one entity or action from another entity or action without necessarily requiring or implying any actual such relationship or order between such entities or actions. Also, the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other identical elements in a process, method, article, or apparatus that comprises the element.
The above description is merely exemplary of the present application and is presented to enable those skilled in the art to understand and practice the present application. Various modifications to these embodiments will be readily apparent to those skilled in the art, and the generic principles defined herein may be applied to other embodiments without departing from the spirit or scope of the application. Thus, the present application is not intended to be limited to the embodiments shown herein but is to be accorded the widest scope consistent with the principles and novel features disclosed herein.
Claims (10)
1. A model building system, the system comprising: the system comprises a data acquisition module, a data detection module, a feature extraction module, a feature analysis module, a feature processing module and a model training module;
the data acquisition module is used for acquiring sample data of at least one data source;
the data detection module is used for detecting the sample data of at least one data source based on a preset quality standard corresponding to the data source and taking the sample data meeting the preset quality standard as target sample data;
the characteristic extraction module is used for extracting the data characteristics of the target sample data;
the characteristic analysis module is used for selecting the data characteristics which accord with preset characteristic indexes from the data characteristics as target characteristics;
the characteristic processing module is used for determining the distribution characteristics of the target characteristics and performing box separation processing on the target characteristics based on the distribution characteristics to obtain a plurality of characteristic boxes;
and the model training module is used for training a target model based on a plurality of feature boxes.
2. The system of claim 1, further comprising: a model analysis module;
and the model analysis module is used for analyzing the execution condition of each module based on the analysis index corresponding to the module so as to obtain the analysis result corresponding to the module.
3. The system of claim 2, further comprising: a report generation module;
the report generation module is used for generating an analysis report based on the analysis result and a preset report format.
4. The system of claim 1, further comprising: and the model deployment module is used for determining a configuration file of the target model and performing online deployment on the target model based on the configuration file.
5. The system of claim 1, wherein the feature extraction module comprises: extracting submodules and deriving the submodules;
the extraction submodule is used for extracting basic features of the target sample data and taking the basic features as the data features;
and the derivative submodule is used for processing the basic characteristics to obtain derivative characteristics, and taking the derivative characteristics as the data characteristics.
6. A method of model construction, the method comprising:
collecting sample data of at least one data source;
detecting sample data of at least one data source based on a preset quality standard corresponding to the data source, and taking the sample data meeting the preset quality standard as target sample data;
extracting data characteristics of the target sample data, and selecting the data characteristics which accord with preset characteristic indexes from the data characteristics as target characteristics;
determining the distribution characteristics of the target characteristics, and performing box separation processing on the target characteristics based on the distribution characteristics to obtain a plurality of characteristic boxes;
training a target model based on a plurality of the feature boxes until the target model converges.
7. The method of claim 6, further comprising:
and aiming at each module, analyzing the execution condition of the module based on the analysis index corresponding to the module to obtain the analysis result corresponding to the module.
8. The method of claim 7, further comprising:
and generating an analysis report based on the analysis result and a preset report format.
9. The method of claim 6, further comprising:
determining a configuration file of the target model;
and performing online deployment on the target model based on the configuration file.
10. The method of claim 6, wherein said extracting data features of said target sample data comprises:
extracting basic features of the target sample data, and taking the basic features as the data features;
and processing the basic features to obtain derivative features, and taking the derivative features as the data features.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110931700.2A CN113657481A (en) | 2021-08-13 | 2021-08-13 | Model construction system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110931700.2A CN113657481A (en) | 2021-08-13 | 2021-08-13 | Model construction system and method |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN113657481A true CN113657481A (en) | 2021-11-16 |
Family
ID=78479846
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110931700.2A Pending CN113657481A (en) | 2021-08-13 | 2021-08-13 | Model construction system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113657481A (en) |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110009479A (en) * | 2019-03-01 | 2019-07-12 | 百融金融信息服务股份有限公司 | Credit assessment method and device, storage medium, computer equipment |
| CN111178690A (en) * | 2019-12-11 | 2020-05-19 | 国网重庆市电力公司北碚供电分公司 | Electricity stealing risk assessment method for electricity consumers based on wind control scoring card model |
| CN111311400A (en) * | 2020-03-30 | 2020-06-19 | 百维金科(上海)信息科技有限公司 | Modeling method and system of grading card model based on GBDT algorithm |
| WO2020143233A1 (en) * | 2019-01-07 | 2020-07-16 | 平安科技（深圳）有限公司 | Method and device for building scorecard model, computer apparatus and storage medium |
| CN112328657A (en) * | 2020-11-03 | 2021-02-05 | 中国平安人寿保险股份有限公司 | Feature derivation method, feature derivation device, computer equipment and medium |
| US20210150415A1 (en) * | 2018-10-24 | 2021-05-20 | Advanced New Technologies Co., Ltd. | Feature selection method, device and apparatus for constructing machine learning model |
-
2021
- 2021-08-13 CN CN202110931700.2A patent/CN113657481A/en active Pending
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20210150415A1 (en) * | 2018-10-24 | 2021-05-20 | Advanced New Technologies Co., Ltd. | Feature selection method, device and apparatus for constructing machine learning model |
| WO2020143233A1 (en) * | 2019-01-07 | 2020-07-16 | 平安科技（深圳）有限公司 | Method and device for building scorecard model, computer apparatus and storage medium |
| CN110009479A (en) * | 2019-03-01 | 2019-07-12 | 百融金融信息服务股份有限公司 | Credit assessment method and device, storage medium, computer equipment |
| CN111178690A (en) * | 2019-12-11 | 2020-05-19 | 国网重庆市电力公司北碚供电分公司 | Electricity stealing risk assessment method for electricity consumers based on wind control scoring card model |
| CN111311400A (en) * | 2020-03-30 | 2020-06-19 | 百维金科(上海)信息科技有限公司 | Modeling method and system of grading card model based on GBDT algorithm |
| CN112328657A (en) * | 2020-11-03 | 2021-02-05 | 中国平安人寿保险股份有限公司 | Feature derivation method, feature derivation device, computer equipment and medium |
Non-Patent Citations (1)
| Title |
|---|
| 高昊阳: "基于大数据的P2P金融风险控制***的设计与实现", 《中国优秀硕士学位论文全文数据库 信息科技辑》, no. 1, pages 138 - 1876 * |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN110147726B (en) | Service quality inspection method and device, storage medium and electronic device | |
| CN108256074B (en) | Verification processing method and device, electronic equipment and storage medium | |
| CN111274814B (en) | Novel semi-supervised text entity information extraction method | |
| EP4016375A1 (en) | Video classification method, device and system | |
| CN112348519A (en) | Method and device for identifying fraudulent user and electronic equipment | |
| CN105469219A (en) | Method for processing power load data based on decision tree | |
| CN114861120B (en) | Flotation froth grade calculation method, device, electronic equipment and medium | |
| CN116541911A (en) | Packaging design system based on artificial intelligence | |
| CN114492926A (en) | Method and system for analyzing and predicting potential safety hazard text of coal mine | |
| CN110310012B (en) | Data analysis method, device, equipment and computer readable storage medium | |
| CN116756688A (en) | Public opinion risk discovery method based on multi-mode fusion algorithm | |
| CN111708810A (en) | Model optimization recommendation method and device and computer storage medium | |
| US8341170B2 (en) | Apparatus and method for visualizing technology change | |
| CN116522912B (en) | Training method, device, medium and equipment for package design language model | |
| CN112800232A (en) | Big data based case automatic classification and optimization method and training set correction method | |
| CN113657481A (en) | Model construction system and method | |
| CN117132383A (en) | Credit data processing method, device, equipment and readable storage medium | |
| CN116522171A (en) | Electric power field fault analysis method and system based on big data | |
| CN115277205B (en) | Model training method and device and port risk identification method | |
| CN115358473A (en) | Power load prediction method and prediction system based on deep learning | |
| CN114139643B (en) | Monoglyceride quality detection method and system based on machine vision | |
| CN112613533B (en) | Image segmentation quality evaluation network system and method based on ordering constraint | |
| CN111274404B (en) | Small sample entity multi-field classification method based on man-machine cooperation | |
| CN117076454B (en) | Engineering quality acceptance form data structured storage method and system | |
| CN116911852B (en) | RPA user dynamic information monitoring method and system |
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 |