Discriminative feature generation for classification of imbalanced data

TL;DR

This paper introduces a supervised discriminative feature generation method using a modified GAN structure to improve classification performance on imbalanced datasets by augmenting minority class features with attention mechanisms.

Contribution

The paper presents a novel DFG approach that enhances minority class feature augmentation through a four-network GAN and supervised attention, improving classification in imbalanced data scenarios.

Findings

Significant improvement in classification accuracy on imbalanced datasets.

Enhanced feature diversity and label preservation through DFG.

Effective use of attention mechanisms for discriminative feature augmentation.

Abstract

The data imbalance problem is a frequent bottleneck in the classification performance of neural networks. In this paper, we propose a novel supervised discriminative feature generation (DFG) method for a minority class dataset. DFG is based on the modified structure of a generative adversarial network consisting of four independent networks: generator, discriminator, feature extractor, and classifier. To augment the selected discriminative features of the minority class data by adopting an attention mechanism, the generator for the class-imbalanced target task is trained, and the feature extractor and classifier are regularized using the pre-trained features from a large source data. The experimental results show that the DFG generator enhances the augmentation of the label-preserved and diverse features, and the classification results are significantly improved on the target task. The feature generation model can contribute greatly to the development of data augmentation methods through discriminative feature generation and supervised attention methods.