Method, apparatus and device for generating model and storage medium

An electronic device, comprising:

• at least one processor;
• and
• a memory storing instructions that, when executed by the at least one processor, cause the at least one processor to perform operations for generating a model, the operations comprising:
• acquiring sample resource features and sample labels corresponding to the sample resource features, wherein the sample resource features comprise one of sample image features, sample text features, or sample voice features;
• training an eXtreme Gradient Boosting, XGBoost, model of a decision tree by taking the sample resource features as an input of the XGBoost model, and taking the sample labels corresponding to the sample resource features as a desired output, to obtain the trained XGBoost model;
• determining a first screening factor based on parameters associated with the trained XGBoost model, the parameters associated with the trained XGBoost model comprising coverage and a correlation coefficient, the determining comprising setting a weight of a coverage and a weight of a correlation coefficient according to feature screening requirements, and performing a weighted sum of the coverage and the correlation coefficient to obtain the first screening factor, wherein the correlation coefficient is a correlation coefficient between the sample resource features and the sample labels corresponding to the sample resource features, and the coverage is a ratio of a difference between a total number of samples and a number of samples with missed features to the total number of samples, wherein the total number of samples is a number of all samples involved in a process of training the XGBoost model, and the number of samples with missed features is a number of samples, which have missed features in all samples;
• determining first resource features from the sample resource features according to the first screening factor;
• training a logistic regression, LR, model, by taking the first resource features as an input of the LR model, and taking sample labels corresponding to the first resource features as a desired output to obtain a trained LR model;
• determining a second screening factor according to parameters associated with the trained LR model, and determining second resource features from the first resource features based on the second screening factor, and obtaining features of a target model based on the second resource features;
• training a machine learning model, by taking the features of the target model as an input of the target model, and taking the sample labels corresponding to the features of the target model as a target value of the target model, to obtain the trained target model;
• and
• predicting the sample image features, sample text features, or sample voice features using the trained target model to obtain a binary classification result.

A non-transitory computer-readable storage medium storing computer instructions that, when executed by a computer, cause the computer to perform operations for generating a model, the operations comprising:

• acquiring sample resource features and sample labels corresponding to the sample resource features, wherein the sample resource features comprise one of sample image features, sample text features, or sample voice features;
• training an eXtreme Gradient Boosting, XGBoost, model of a decision tree by taking the sample resource features as an input of the XGBoost model, and taking the sample labels corresponding to the sample resource features as a desired output, to obtain the trained XGBoost model;
• determining a first screening factor based on meters associated the trained XGBoost model, the parameter associated with the trained XGBoost model comprising coverage and a correlation coefficient, the determining comprising:
• setting a weight of a coverage and a weight of a correlation coefficient according to feature screening requirements, and performing a weighted sum of the coverage and the correlation coefficient to obtain the first screening factor, wherein the correlation coefficient is a correlation coefficient between the sample resource features and the sample labels corresponding to the sample resource features, and the coverage is a ratio of a difference between a total number of samples and a number of samples with missed features to the total number of samples, wherein the total number of samples is a number of all samples involved in a process of training the XGBoost model, and the number of samples with missed features is a number of samples, which have missed features in all samples;
• determining first resource features from the sample resource features according to the first screening factor;
• training a logistic regression, LR, model, by taking the first resource features as an input of the LR model, and taking sample labels corresponding to the first resource features as a desired output to obtain a trained LR model;
• determining a second screening factor according to parameters associated with the trained LR model, and determining second resource features from the first resource features based on the second screening factor, and obtaining features of a target model based on the second resource features;
• training a machine learning model, by taking the features of the target model as an input of the target model, and taking the sample labels corresponding to the features of the target model as a target value of the target model, to obtain the trained target model;
• and
• predicting the sample image features, sample text features, or sample voice features using the trained target model to obtain a binary classification result.