Transfer learning method and device

Abstract

The invention discloses a transfer learning method and device. The method comprises setting relevant parameters of transfer learning and performing initialization; starting transfer learning iteration to obtain automatic annotation data; when the number of the iteration meets an iteration cycle, performing error detection on the automatic annotation data serving as samples in the iteration cycle to determine the sample relative quality of the iteration cycle; according to the sample relative quality, determining whether to delete or retain the samples and determining whether to continue the transfer learning iteration; when the transfer learning iteration is stopped, outputting the retainer samples and a transfer classifier. The transfer learning method has the advantage that, during a transfer learning process, the learning process is divided according to the transfer cycle, error detection is performed every full cycle, the samples are screened according to the sample relative quality determined through the error detection, so that low-quality samples can be eliminated to achieve the aim of improving the sample quality of automatic annotation data during the transfer learning process and further to improve the accuracy of a system applying the transfer learning method.

Description

technical field [0001] The present invention relates to the field of machine learning, in particular to a transfer learning method and device. Background technique [0002] In traditional classification learning, in order to ensure the accuracy and high reliability of the trained classification model, there are two basic assumptions: (1) the training samples used for learning and the new test samples satisfy the condition of independent and identical distribution; (2) There must be enough available training samples to learn a good classification model. However, in practical applications, it is found that these two conditions are often not satisfied. However, transfer learning, an important branch of machine learning, relaxes two basic assumptions in traditional machine learning. Migration learning is mainly aimed at obtaining knowledge training related models from the resource-rich source domain and the target domain, and then solving the problem of the target domain with ...

AI Technical Summary

Problems solved by technology

In the early stage of the entire learning process, due to the expansion of the training set by automatically labeled samples, the performance index of the algorithm will increase, but with the accumulation of misclassified samples, the quality of the samples will deteriorate, and the performance of the algorithm will decline in the middle and late stages of the learning process.

For practical tasks with unknown answers, this problem makes it difficult for experimenters to determine the time point when the algorithm stops iterating, which affects the performance that the system can achieve

[0005] The current machine learning methods for transfer learning often focus on the accuracy of the baseline method, and the problem of wrongly labeled samples during the transfer process has not been solved.

Images