Sample classification method based on active learning and neural network

Abstract

The invention discloses a sample classification method based on active learning and a neural network, and belongs to the field of machine learning in intelligent science and technology. According to the method, the uncertainty of a neural network model to sample points is taken as a reference, three traditional uncertainty indexes of Least content, Margin and Entry are calculated respectively, the three indexes are used for voting samples, the sample with the highest vote number is the finally screened sample points, the sample points are the most uncertain samples of the model, and then training of the neural network model is facilitated. The number of sample points needing to be marked can be effectively reduced, the marking cost is reduced, and the classification precision of the model is improved.

Description

technical field [0001] The invention belongs to the technical field of machine learning in intelligence science and technology. Specifically, it involves a method based on neural network, introducing active learning and using three types of uncertainty indicators to vote samples, manually marking samples with high votes, constructing high-quality training sample sets, and further improving the classification accuracy of neural networks. Background technique [0002] The learning of traditional neural network models is mainly based on supervised learning. This learning method requires a large number of labeled samples to guide the model for training, so as to continuously improve the performance of the model in the target field, and finally obtain a certain generalization ability and put it into practical application scenarios. middle. However, the labeling of training samples is generally done by experts in related fields, and obtaining high-quality labels requires a lot of...

AI Technical Summary

Problems solved by technology

[0005] The above three uncertain screening strategies each represent a part of the uncertainty of the label of the model to the sample, and cannot completely measure the uncertainty of the model to the sample. In practical applications, the classification accuracy of the model is relatively low.

Images