FREQuency ATTribution: benchmarking frequency-based occlusion for time series data

TL;DR

This paper introduces FreqAtt, a frequency-based interpretability framework for time-series neural networks, demonstrating its robustness and effectiveness in highlighting relevant input features compared to existing methods.

Contribution

The paper presents FreqAtt, a novel frequency domain analysis method for post-hoc interpretation of time-series neural networks, improving relevance detection and robustness.

Findings

Frequency-based attribution outperforms traditional methods in relevance detection.

Combining frequency analysis with traditional attribution enhances interpretability.

FreqAtt is robust to signal fluctuations and provides comprehensive insights.

Abstract

Deep neural networks are among the most successful algorithms in terms of performance and scalability across different domains. However, since these networks are black boxes, their usability is severely restricted due to a lack of interpretability. Existing interpretability methods do not address the analysis of time-series-based networks specifically enough. This paper shows that an analysis in the frequency domain can not only highlight relevant areas in the input signal better than existing methods but is also more robust to fluctuations in the signal. In this paper, FreqAtt is presented - a framework that enables post-hoc interpretation of time-series analysis. To achieve this, the relevant frequencies are evaluated, and the signal is either filtered or the relevant input data is marked. FreqAtt is evaluated using a wide range of statistical metrics to provide a broad overview of its performance. The results show that using frequency-based attribution, especially in combination with traditional attribution on top of the frequency-optimized signal, provides strong performance across different metrics.