Which products activate a product? An explainable machine learning approach
This work addresses the explainability gap in policy recommendations for export strategies, providing a domain-specific tool for economists and policymakers.
The paper tackles the problem of interpreting tree-based machine learning predictions for export feasibility by proposing a procedure to statistically validate product importance, identifying 'explainers' that significantly increase the probability of exporting a target product, with results showing a positive correlation between product complexity and explainer complexity.
Tree-based machine learning algorithms provide the most precise assessment of the feasibility for a country to export a target product given its export basket. However, the high number of parameters involved prevents a straightforward interpretation of the results and, in turn, the explainability of policy indications. In this paper, we propose a procedure to statistically validate the importance of the products used in the feasibility assessment. In this way, we are able to identify which products, called explainers, significantly increase the probability to export a target product in the near future. The explainers naturally identify a low dimensional representation, the Feature Importance Product Space, that enhances the interpretability of the recommendations and provides out-of-sample forecasts of the export baskets of countries. Interestingly, we detect a positive correlation between the complexity of a product and the complexity of its explainers.