FreqRISE: Explaining time series using frequency masking
This work addresses the need for explainable AI in critical domains like healthcare and finance, offering a novel approach to time-series explanation.
The paper tackled the problem of explaining time-series data by proposing that salient information is often localized in the frequency domain, and introduced FreqRISE, a method using frequency masking that outperforms strong baselines across multiple tasks.
Time-series data are fundamentally important for many critical domains such as healthcare, finance, and climate, where explainable models are necessary for safe automated decision making. To develop explainable artificial intelligence in these domains therefore implies explaining salient information in the time series. Current methods for obtaining saliency maps assume localized information in the raw input space. In this paper, we argue that the salient information of a number of time series is more likely to be localized in the frequency domain. We propose FreqRISE, which uses masking-based methods to produce explanations in the frequency and time-frequency domain, and outperforms strong baselines across a number of tasks. The source code is available here: \url{https://github.com/theabrusch/FreqRISE}.