Noise label correction method

Abstract

The invention provides a noise label relabeling method, which comprises the following steps of: 1, classifying observation samples by using a base classifier, estimating a noise rate, and identifyingnoise label data; and step 2, re-labeling the noise label sample by using a base classifier to obtain a corrected clean sample data set of the noise label sample.

Description

technical field [0001] The invention relates to a data mining technology, in particular to a noise label correction method. Background technique [0002] Traditional supervised learning classification problems usually assume that the labels of the dataset are complete, that is, there is a noise-free correct label for each dataset sample. However, in the real world, due to the randomness of the labeling process, sample labels are easily polluted by noise, resulting in inaccurate sample labels. The generation of noisy data is usually related to the acquisition method of the data set. For example, in the process of labeling the original data, the amount of sample data provided to the labeler is not enough to cause the labeler to misclassify the sample, or because the classification process itself is a subjective process or the professional knowledge of the labeler is not enough to ensure the correct classification sex. Various popular data labeling platforms are also one of ...

AI Technical Summary

Problems solved by technology

Natarajan proposed two methods to modify the loss function. The first method constructs an unbiased estimator of the correct distribution from the noise distribution, but the estimator may still be a non-convex function even if the original loss function is a convex function.

Images