Towards Explainable Meta-learning

Meta-learning is a field that aims at discovering how different machine learning algorithms perform on a wide range of predictive tasks. Such knowledge speeds up the hyperparameter tuning or feature engineering. With the use of surrogate models, various aspects of the predictive task such as meta-features, landmarker models, etc., are used to predict expected performance. State-of-the-art approaches focus on searching for the best meta-model but do not explain how these different aspects contribute to its performance. However, to build a new generation of meta-models, we need a deeper understanding of the importance and effect of meta-features on model tunability. This paper proposes techniques developed for eXplainable Artificial Intelligence (XAI) to examine and extract knowledge from black-box surrogate models. To our knowledge, this is the first paper that shows how post-hoc explainability can be used to improve meta-learning.

Automated summary

Meta-learning is a field that explores the performance of different machine learning algorithms on predictive tasks and aims to accelerate hyperparameter tuning or feature engineering. Current research focuses on finding the best meta-model without explaining how different aspects contribute to its performance. To build new generation of meta-models, a deeper understanding of the importance and effect of meta-features on model tunability is needed.