FIC-TSC: Learning Time Series Classification with Fisher Information Constraint

TL;DR

FIC-TSC introduces a Fisher information constraint-based training framework for time series classification, improving generalization under domain shifts by guiding models toward flatter minima, validated on extensive datasets.

Contribution

The paper proposes a novel Fisher information constraint approach for time series classification, enhancing robustness to domain shifts and outperforming recent methods.

Findings

Outperforms 14 state-of-the-art methods on multiple datasets

Guides models toward flatter minima for better generalization

Effective in handling distribution shifts in time series data

Abstract

Analyzing time series data is crucial to a wide spectrum of applications, including economics, online marketplaces, and human healthcare. In particular, time series classification plays an indispensable role in segmenting different phases in stock markets, predicting customer behavior, and classifying worker actions and engagement levels. These aspects contribute significantly to the advancement of automated decision-making and system optimization in real-world applications. However, there is a large consensus that time series data often suffers from domain shifts between training and test sets, which dramatically degrades the classification performance. Despite the success of (reversible) instance normalization in handling the domain shifts for time series regression tasks, its performance in classification is unsatisfactory. In this paper, we propose \textit{FIC-TSC}, a training framework for time series classification that leverages Fisher information as the constraint. We theoretically and empirically show this is an efficient and effective solution to guide the model converge toward flatter minima, which enhances its generalizability to distribution shifts. We rigorously evaluate our method on 30 UEA multivariate and 85 UCR univariate datasets. Our empirical results demonstrate the superiority of the proposed method over 14 recent state-of-the-art methods.