Synthetic time series for multi series demand forecasting

<pre>Global feature importance methods are one of the core tools for interpreting the role of explanatory variables in machine learning models,<br>however, using them more complex forecasting tasks involving multiple time series can be challenging.<br>This study focuses on tree-based ensemble models, applied to multi-series product demand forecasting.<br>To evaluate the different global feature importance methods, we generate simulated datasets with controlled dependencies on<br>lagged values and external demand drivers.<br>We compare model-specific and model-agnostic global importance methods, including SHAP values, permutation importance, and tree-specific gain- and split-based importance.<br>Our analysis focuses on uncovering pitfalls in applying these methods, including problems introduced by auto-correlation and feature scaling.</pre>