The landscape of empirical risk for nonconvex losses
Annals of Statistics, pp. 2747-2774, 2018.
Most high-dimensional estimation and prediction methods propose to minimize a cost function (empirical risk) that is written as a sum of losses associated to each data point. this paper we focus on the case of non-convex losses, which is practically important but still poorly understood. Classical empirical process theory implies uniform...More
PPT (Upload PPT)