Class-incremental learning on multivariate time series via shape-aligned temporal distillation

• Main Authors: Qiao, Zhongzheng, Hu, Minghui, Jiang, Xudong, Suganthan, Ponnuthurai Nagaratnam, Savitha, Ramasamy
• 其他作者: School of Electrical and Electronic Engineering
• 格式: Conference or Workshop Item
• 語言: English
• 出版: 2023
• 主題: Engineering::Computer science and engineering::Computing methodologies::Artificial intelligence; Continual Learning; Multivariate Time Series Classification; Knowledge Distillation; Dynamic Time Warping
• 在線閱讀: https://hdl.handle.net/10356/165392

Class-incremental learning (CIL) on multivariate time series (MTS) is an important yet understudied problem. Based on practical privacy-sensitive circumstances, we propose a novel distillation-based strategy using a single-headed classifier without saving historical samples. We propose to exploit So...