LoMEF: A Framework to Produce Local Explanations for Global Model Time Series Forecasts
This addresses the problem of reduced trust and confidence for stakeholders in time series forecasting by providing interpretable explanations for black-box global models.
The paper tackles the lack of interpretability in Global Forecasting Models (GFMs) by proposing LoMEF, a local model-agnostic framework that uses simpler univariate surrogate models to explain GFM forecasts, showing benefits in accuracy, fidelity, stability, and comprehensibility.
Global Forecasting Models (GFM) that are trained across a set of multiple time series have shown superior results in many forecasting competitions and real-world applications compared with univariate forecasting approaches. One aspect of the popularity of statistical forecasting models such as ETS and ARIMA is their relative simplicity and interpretability (in terms of relevant lags, trend, seasonality, and others), while GFMs typically lack interpretability, especially towards particular time series. This reduces the trust and confidence of the stakeholders when making decisions based on the forecasts without being able to understand the predictions. To mitigate this problem, in this work, we propose a novel local model-agnostic interpretability approach to explain the forecasts from GFMs. We train simpler univariate surrogate models that are considered interpretable (e.g., ETS) on the predictions of the GFM on samples within a neighbourhood that we obtain through bootstrapping or straightforwardly as the one-step-ahead global black-box model forecasts of the time series which needs to be explained. After, we evaluate the explanations for the forecasts of the global models in both qualitative and quantitative aspects such as accuracy, fidelity, stability and comprehensibility, and are able to show the benefits of our approach.