Stochastic Low-Rank Bandits

12/13/2017
by   Branislav Kveton, et al.
0

Many problems in computer vision and recommender systems involve low-rank matrices. In this work, we study the problem of finding the maximum entry of a stochastic low-rank matrix from sequential observations. At each step, a learning agent chooses pairs of row and column arms, and receives the noisy product of their latent values as a reward. The main challenge is that the latent values are unobserved. We identify a class of non-negative matrices whose maximum entry can be found statistically efficiently and propose an algorithm for finding them, which we call LowRankElim. We derive a O((K + L) (d) Δ^-1 n) upper bound on its n-step regret, where K is the number of rows, L is the number of columns, d is the rank of the matrix, and Δ is the minimum gap. The bound depends on other problem-specific constants that clearly do not depend K L. To the best of our knowledge, this is the first such result in the literature.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
08/10/2016

Stochastic Rank-1 Bandits

We propose stochastic rank-1 bandits, a class of online learning problem...
research
05/21/2017

Nice latent variable models have log-rank

Matrices of low rank are pervasive in big data, appearing in recommender...
research
04/29/2020

Quantum and approximation algorithms for maximum witnesses of Boolean matrix products

The problem of finding maximum (or minimum) witnesses of the Boolean pro...
research
03/19/2017

Bernoulli Rank-1 Bandits for Click Feedback

The probability that a user will click a search result depends both on i...
research
06/26/2018

Main effects and interactions in mixed and incomplete data frames

A mixed data frame (MDF) is a table collecting categorical, numerical an...
research
12/01/2013

Stochastic continuum armed bandit problem of few linear parameters in high dimensions

We consider a stochastic continuum armed bandit problem where the arms a...
research
12/07/2017

Sets of Stochastic Matrices with Converging Products: Bounds and Complexity

An SIA matrix is a stochastic matrix whose sequence of powers converges ...

Please sign up or login with your details

Forgot password? Click here to reset