Towards Healing the Blindness of Score Matching
This addresses a specific issue in statistical estimation for researchers, but appears incremental as it builds on existing divergence methods.
The paper tackled the blindness problem of score-based divergences in multi-modal distributions by proposing a new family of divergences, reporting improved performance in density estimation compared to traditional approaches.
Score-based divergences have been widely used in machine learning and statistics applications. Despite their empirical success, a blindness problem has been observed when using these for multi-modal distributions. In this work, we discuss the blindness problem and propose a new family of divergences that can mitigate the blindness problem. We illustrate our proposed divergence in the context of density estimation and report improved performance compared to traditional approaches.