ITS2Graph: Graph-based generative adversarial learning for imbalanced time series classification

Chang Liu

doi:10.1016/j.neunet.2025.107770

ITS2Graph: Graph-based generative adversarial learning for imbalanced time series classification

2025-06-28

0

PRE

AI

DOI：10.1016/j.neunet.2025.107770

Original

Original request for help

Save

Abstract

En 中文

• We propose a novel graph-based method for imbalanced time series classification. • Our method converts the time series classification problem into the node classification problem within a graph. • Our proposed graph construction method effectively captures higher-order correlation patterns between time series. • We utilize a graph generator to synthesize the attributes of minority class nodes and the network topology.

ITS2Graph: Graph-based generative adversarial learning for imbalanced time series classification

Abstract

Journal

Organization