This paper introduces a new method using dyadic decision trees for estimating a classification or a regression function in a multiclass classification problem. The estimator is bas...
We present a unified framework for reasoning about worst-case regret bounds for learning algorithms. This framework is based on the theory of duality of convex functions. It brin...
We propose Shotgun, a parallel coordinate descent algorithm for minimizing L1regularized losses. Though coordinate descent seems inherently sequential, we prove convergence bounds...
Joseph K. Bradley, Aapo Kyrola, Danny Bickson, Car...
This paper addresses the important tradeoff between privacy and learnability, when designing algorithms for learning from private databases. We focus on privacy-preserving logisti...
Convex optimization problems arising in applications, possibly as approximations of intractable problems, are often structured and large scale. When the data are noisy, it is of i...