Research Article
D-AE: A Discriminant Encode-Decode Nets for Data Generation
@INPROCEEDINGS{10.1007/978-3-031-54528-3_6, author={Gongju Wang and Yulun Song and Yang Li and Mingjian Ni and Long Yan and Bowen Hu and Quanda Wang and Yixuan Li and Xingru Huang}, title={D-AE: A Discriminant Encode-Decode Nets for Data Generation}, proceedings={Collaborative Computing: Networking, Applications and Worksharing. 19th EAI International Conference, CollaborateCom 2023, Corfu Island, Greece, October 4-6, 2023, Proceedings, Part II}, proceedings_a={COLLABORATECOM PART 2}, year={2024}, month={2}, keywords={Imbalanced datasets Discriminant-Autoencoder (D-AE) Discriminant- loss function}, doi={10.1007/978-3-031-54528-3_6} }- Gongju Wang
Yulun Song
Yang Li
Mingjian Ni
Long Yan
Bowen Hu
Quanda Wang
Yixuan Li
Xingru Huang
Year: 2024
D-AE: A Discriminant Encode-Decode Nets for Data Generation
COLLABORATECOM PART 2
Springer
DOI: 10.1007/978-3-031-54528-3_6
Abstract
Imbalanced datasets often result in poor predictive model performance. To address this, minority class sample expansion is used, but two challenges remain. The first is to use algorithms to learn the main features of minority class samples, and the second is to differentiate the generated data from the majority class samples. To tackle these challenges in binary classification, we propose the Discriminant-Autoencoder (D-AE) algorithm. It has two mechanisms based on our insights. Firstly, an autoencoder is used to learn the main features of minority class samples by reconstructing the data with added noise. Secondly, a discriminator is trained on the raw data to distinguish the generated data from the majority class samples. Our proposed loss function, Discriminant-(L_\theta ), balances the discriminant and reconstruction losses. Results from experiments on three datasets show that D-AE outperforms baseline algorithms and improves dataset applicability.
