Motivated by online retail applications, we study the online personalized assortment optimization problem. A seller conducts sales by offering assortments of products to a stream of arriving customers. The customers’ purchase behaviour follows their respective personalized Multinomial Logit choice models, which vary according to their individual attributes. The seller aims to maximize his revenue by offering personalized assortments to the customers, notwithstanding his uncertainty about the customers’ choice models. We propose a Thompson Sampling based policy, policy Pao-Ts, where surrogate models for the latent choice models are constructed using samples from a progressively updated posterior distribution. We derive bounds on the revenue loss, namely Bayesian regret, incurred by policy Pao-Ts, in comparison to the optimal policy which is provided with the latent models. The regret bounds hold even when the customers’ attributes vary arbitrarily, but not independently and identically distributed.

Speaker Bio

Cheung Wang Chi is currently a research scientist at the Institute of High Performance Computing in the Agency of Science, Technology and Research (ASTAR). Wang Chi has completed his PhD in the MIT Operations Center, advised by David Simchi-Levi. He is interested in data driven optimization, with applications to revenue management and inventory control models. He is the recipient of the ASTAR NSS (BS) scholarship from 2007 to 2010, as well as the ASTAR NSS (PhD) scholarship from 2011 to 2016. He is also a finalist in the George Nicholson Student Paper Competition in 2015.

For more information about the ESD Seminars Series, please contact Karthik Natarajan at karthik_natarajan@sutd.edu.sg.