EXCODER: EXplainable Classification Of DiscretE time series Representations

TL;DR

This paper explores transforming time series data into discrete latent representations to improve the explainability of deep learning models, demonstrating that such representations enable more concise, faithful, and computationally efficient explanations without sacrificing classification accuracy.

Contribution

The work introduces a novel approach using discrete latent representations for explainable time series classification and proposes SSA, a new metric for evaluating explanation quality.

Findings

Discrete representations preserve classification performance.

XAI explanations become more concise and structured.

SSA effectively measures explanation faithfulness.

Abstract

Deep learning has significantly improved time series classification, yet the lack of explainability in these models remains a major challenge. While Explainable AI (XAI) techniques aim to make model decisions more transparent, their effectiveness is often hindered by the high dimensionality and noise present in raw time series data. In this work, we investigate whether transforming time series into discrete latent representations-using methods such as Vector Quantized Variational Autoencoders (VQ-VAE) and Discrete Variational Autoencoders (DVAE)-not only preserves but enhances explainability by reducing redundancy and focusing on the most informative patterns. We show that applying XAI methods to these compressed representations leads to concise and structured explanations that maintain faithfulness without sacrificing classification performance. Additionally, we propose Similar Subsequence Accuracy (SSA), a novel metric that quantitatively assesses the alignment between XAI-identified salient subsequences and the label distribution in the training data. SSA provides a systematic way to validate whether the features highlighted by XAI methods are truly representative of the learned classification patterns. Our findings demonstrate that discrete latent representations not only retain the essential characteristics needed for classification but also offer a pathway to more compact, interpretable, and computationally efficient explanations in time series analysis.