Locally Adaptive and Differentiable Regression

Mingxuan Han, Varun Shankar, Jeff M Phillips +1

Over-parameterized models like deep nets and random forests have become very popular in machine learning. However, the natural goals of continuity and differentiability, common in…