Learning Supervised PageRank with Gradient-Based and Gradient-Free Optimization Methods

Part of Advances in Neural Information Processing Systems 29 (NIPS 2016)

Bibtex Metadata Paper Reviews Supplemental

Authors

Lev Bogolubsky, Pavel Dvurechenskii, Alexander Gasnikov, Gleb Gusev, Yurii Nesterov, Andrei M. Raigorodskii, Aleksey Tikhonov, Maksim Zhukovskii

Abstract

In this paper, we consider a non-convex loss-minimization problem of learning Supervised PageRank models, which can account for features of nodes and edges. We propose gradient-based and random gradient-free methods to solve this problem. Our algorithms are based on the concept of an inexact oracle and unlike the state-of-the-art gradient-based method we manage to provide theoretically the convergence rate guarantees for both of them. Finally, we compare the performance of the proposed optimization methods with the state of the art applied to a ranking task.