Active Learning for Non-Parametric Choice Models

  title={Active Learning for Non-Parametric Choice Models},
  author={Fransisca Susan and Negin Golrezaei and Ehsan Emamjomeh-Zadeh and David Kempe},
We study the problem of actively learning a non-parametric choice model based on consumers’ decisions. We present a negative result showing that such choice models may not be identifiable. To overcome the identifiability problem, we introduce a directed acyclic graph (DAG) representation of the choice model, which in a sense captures as much information about the choice model as could information-theoretically be identified. We then consider the problem of learning an approximation to this DAG… 

Figures and Tables from this paper



A Market Discovery Algorithm to Estimate a General Class of Nonparametric Choice Models

A market discovery algorithm that starts with a parsimonious set of types and enlarge it by automatically generating new types that increase the likelihood value is proposed, which improves the root mean square errors between predicted and observed purchases computed under independent demand model estimates.

MNL-Bandit: A Dynamic Learning Approach to Assortment Selection

An efficient algorithm is given that simultaneously explores and exploits, achieving performance independent of the underlying parameters, and is adaptive in the sense that its performance is near-optimal in both the "well separated" case, as well as the general parameter setting where this separation need not hold.

A Nonparametric Approach to Modeling Choice with Limited Data

A non-parametric approach in which the data automatically selects the 'right' choice model for revenue predictions, which demonstrates a 20% improvement in prediction accuracy over state-of-the art benchmark models.

Technical Note - An Expectation-Maximization Method to Estimate a Rank-Based Choice Model of Demand

The EM method is applied to jointly estimate the arrival rate of customers and the pmf of the rank-based choice model, and it is shown that it leads to a simple and highly ecient estimation procedure.

A Nonparametric Joint Assortment and Price Choice Model

A tractable nonparametric expectation maximization (EM) algorithm is developed to fit the model to the aggregate transaction data and an efficient algorithm is designed to determine the profit-maximizing combination of offer set and price.

An Optimal Greedy Heuristic with Minimal Learning Regret for the Markov Chain Choice Model

A greedy heuristic is developed that finds an optimal assortment for the Markov Chain model and runs in $O(n^2)$ iterations and has performance bound $1/n$ for any regular choice model which is best possible among polynomial heuristics.

Optimal Algorithms for Assortment Selection Under Ranking-Based Consumer Choice Models

A ranking-based consumer choice model is used to represent consumer preferences: every customer has a ranking of the potential products in the category and purchases his highest ranked product (if any) offered in the assortment.

Robust Assortment Optimization in Revenue Management Under the Multinomial Logit Choice Model

This work gives a complete characterization of the optimal policy in both settings, shows that it can be computed efficiently, and derive operational insights, and proposes a family of uncertainty sets that enables the decision maker to control the trade-off between increasing the average revenue and protecting against the worst-case scenario.

Revenue Management Under a General Discrete Choice Model of Consumer Behavior

This paper analyses a single-leg reserve management problem in which the buyers' choice behavior is modeled explicitly and develops an estimation procedure based on the expectation-maximization (EM) method that jointly estimates arrival rates and choice model parameters when no-purchase outcomes are unobservable.

Generalized random utility model