Advancements in Rényi entropy and divergence estimation for model assessment

Entropy and divergence, fundamental concepts in machine learning and computer science, have gained significant traction over the past decade. Statisticians have been developing estimators for these measures, advancing computational analysis. In this paper, we present nonparametric estimators for Rényi entropy and divergence. Through a range of examples, we showcase the effectiveness of our approach, demonstrating its applicability across various contexts. Furthermore, we leverage these estimators for model assessment.