Scaling Time Series Classification via XAI-Driven Data Reduction
Machine LearningArtificial Intelligence
Summary
The authors created a method called drXAI that uses explainable AI tools to pick out the most important parts of time series data for classification tasks. This helps reduce the amount of data needed, which is important because powerful models can be very slow or run out of memory with long or complex sequences. They tested drXAI on both made-up and real data, showing it keeps accuracy high while using much less data. Their method also lets big models work on datasets that were too large to handle before. This work shows explainable AI can help not just with understanding models, but also with improving their efficiency.
Authors
Davide Italo Serramazza, Thach Le Nguyen, Georgiana Ifrim
Abstract
Explainable AI (XAI) for time series has seen significant algorithmic growth, but its utility in providing measurable performance gains for downstream tasks remains under-explored. This paper bridges this gap by introducing drXAI, a novel methodology that repurposes XAI attribution methods for effective data reduction in Time Series Classification (TSC). The core challenge in modern TSC is scalability; state-of-the-art models, such as Transformers, exhibit quadratic complexity relative to sequence length and linear complexity relative to the number of channels. This renders them computationally prohibitive for massive datasets. drXAI addresses this by using a fast, GPU-accelerated classifier (Hydra) to generate local attributions. We aggregate these into global feature importance scores and employ an automated elbow-cut heuristic to select the most salient features without requiring manual thresholds. We evaluate our approach on both synthetic and real-world univariate and multivariate datasets. On synthetic benchmarks, drXAI successfully recovers ground-truth features where traditional baselines fail. On real-world data, drXAI achieves between 80% and 90% data reduction while maintaining classification accuracy comparable to models trained on the full dataset. Most importantly, we show that drXAI allows resource-intensive models like ConvTran to scale to datasets that were previously inaccessible due to memory constraints. Our results show the benefits of using XAI not just for interpretability, but as a robust tool for feature selection and scalability in time series analysis. All our code and data are openly available.