Deep transfer learning-based unbalanced classification ensemble method

Abstract

The invention discloses a deep transfer learning-based unbalanced classification ensemble method. The method comprises the following steps that: an auxiliary data set is established; an auxiliary deep network model and a target deep network model are constructed; the auxiliary deep network is trained; the structure and parameters of the auxiliary deep network are transferred to the target deep network; and the products of auprc values are calculated and are adopted as the weights of classifiers, and weighted ensemble is performed on the classification results of each transfer classifier, so that an ensemble classification result is obtained and is adopted as the output of an ensemble classifier. According to the method of the present invention, an improved average precision variance loss function (APE) and an average precision cross-entropy loss function (APCE) are adopted; when the loss cost of samples is calculated, the weights of the samples are dynamically adjusted; and few weights are assigned to majority classification samples, more weights are assigned to minority classification samples, and therefore, the trained deep network attaches more importance to the minority classification samples, and the method is more suitable for the classification of unbalanced data.

Description

technical field [0001] The invention relates to the fields of deep learning, transfer learning and unbalanced classification in machine learning, in particular to an unbalanced classification integration method of deep transfer learning. Background technique [0002] Traditional data classification methods treat different types of data equally and strive to improve the overall classification accuracy. However, in reality, there are many unbalanced data distributions, because some samples are either rare or expensive to collect, making the number of samples of a certain type far less than the number of samples of other types, such as disease detection, bankruptcy prediction, market Customer churn prediction, software defect prediction, etc. The abnormal data class (minority class) in these cases only accounts for 10% to 20% of the normal data class (majority class), and the distribution is unbalanced, even extremely unbalanced. Most of the traditional classification methods...

AI Technical Summary

Problems solved by technology

There are deficiencies in existing deep transfer learning: first, feature transfer may bring benefits to target learning, and at the same time, negative feature transfer may affect target learning; secondly, the choice of deep network structure transfer method makes the learning process take a lot of time and space cost

However, the classifier obtained from the deep network migration learned

Images