Feature Importance for Time Series Data: Improving KernelSHAP
This work addresses the need for explainable AI in time series analysis, offering incremental improvements to existing methods for researchers and practitioners in fields like finance or healthcare.
The paper tackled the problem of applying Shapley value-based feature importance techniques to time series data by presenting closed-form solutions for models like VARMAX and extending KernelSHAP for event detection, resulting in improved interpretability for time series tasks.
Feature importance techniques have enjoyed widespread attention in the explainable AI literature as a means of determining how trained machine learning models make their predictions. We consider Shapley value based approaches to feature importance, applied in the context of time series data. We present closed form solutions for the SHAP values of a number of time series models, including VARMAX. We also show how KernelSHAP can be applied to time series tasks, and how the feature importances that come from this technique can be combined to perform "event detection". Finally, we explore the use of Time Consistent Shapley values for feature importance.