Model training and data classification method and device

Abstract

The invention provides a model training method and device, and relates to the technical field of big data, machine learning and the like. According to the specific implementation scheme, a training sample is obtained according to an auxiliary sample and a target sample which are obtained in advance; a preset weight rule is adopted to generate training weights of the training samples, and the weight rule is used for enabling the weight proportions of the auxiliary samples and the target samples to be the same; executing the following training steps: inputting a training sample and a training weight into a base learning device to obtain the output of the base learning device; and based on the output of the base learner, adopting a migration operator to adjust the training weight until a training stop condition is met, and obtaining a target model. According to the embodiment, the modeling efficiency is improved.

Description

technical field [0001] The present disclosure relates to the field of computer technology, specifically to big data, machine learning and other technical fields, and in particular to a method and device for model training and data classification, electronic equipment, computer-readable storage media, and computer program products. Background technique [0002] When the number of target samples is small, when building a model (for example, a classification model), generally only logistic regression can be used to model, and the modeling effect is not good; while performing EDA (Exploratory Data Analysis, data exploration) on multiple labeled samples When performing analysis, it is necessary to use manual strategies to merge multiple samples to obtain the required training samples, and to perform mixed modeling through training samples. The modeling effect is unstable, and a large amount of manual intervention is required, which takes a long time. Contents of the invention ...

AI Technical Summary

Problems solved by technology

[0002] When the number of target samples is small, when building a model (for example, a classification model), generally only logistic regression can be used to model, and the modeling effect is not good; while performing EDA (Exploratory Data Analysis, data exploration) on multiple labeled samples When analyzing, it is necessary to use a manual strategy to merge multiple samples to obtain the required training samples. The mixed modeling is performed through the training samples. The modeling effect is unstable and requires a lot of manual intervention, which takes a long time.

Images