arXiv subjects

Haotian Jiang

Publications and source records attributed to Haotian Jiang.

57 records · Page 4Linked to original sources

arXiv2017

Practical Algorithms for Best-K Identification in Multi-Armed Bandits

In the Best-$K$ identification problem (Best-$K$-Arm), we are given $N$ stochastic bandit arms with unknown reward distributions. Our goal is to identify the $K$ arms with the largest means with high confidence, by drawing samples from the arms adaptively. This problem is motivated by various practical applications and has attracted considerable attention in the past decade. In this paper, we propose new practical algorithms for the Best-$K$-Arm problem, which have nearly optimal sample complexity bounds (matching the lower bound up to logarithmic factors) and outperform the state-of-the-art algorithms for the Best-$K$-Arm problem (even for $K=1$) in practice.

cs.LG↗

arXiv2016

An Attempt to Design a Better Algorithm for the Uncapacitated Facility Location Problem

The uncapacitated facility location has always been an important problem due to its connection to operational research and infrastructure planning. Byrka obtained an algorithm that is parametrized by $γ$ and proved that it is optimal when $γ>1.6774$. He also proved that the algorithm achieved an approximation ratio of 1.50. A later work by Shi Li achieved an approximation factor of 1.488. In this research, we studied these algorithms and several related works. Although we didn't improve upon the algorithm of Shi Li, our work did provide some insight into the problem. We also reframed the problem as a vector game, which provided a framework to design balanced algorithms for this problem.

cs.DS↗

arXiv2016

Two Stage Optimization with Recourse and Revocation

Two-stage optimization with recourse model is an important and widely used model, which has been studied extensively these years. In this article, we will look at a new variant of it, called the two-stage optimization with recourse and revocation model. This new model differs from the traditional one in that one is allowed to revoke some of his earlier decisions and withdraw part of the earlier costs, which is not unlikely in many real applications, and is therefore considered to be more realistic under many situations. We will adopt several approaches to study this model. In fact, we will develop an LP rounding scheme for some cover problems and show that they can be solved using this scheme and an adaptation of the rounding approach for the deterministic counterpart, provided the polynomial scenario assumption. Stochastic uncapacitated facility location problem will also be studied to show that the approximation algorithm that worked for the two-stage with recourse model worked for this model as well. In addition, we will use other methods to study the model.

cs.DS↗