Social and Information Sciences Laboratory (SISL) Seminar
Learnability and Stochastic Choice
Abstract : In this paper, we study a non-parametric approach to prediction in stochastic choice models in economics. We apply techniques from statistical learning theory to study the problem of learning choice probabilities. A model of stochastic choice is said to be learnable if there exist learning rules defined on choice data that are uniformly consistent. We construct learning rules via the procedure of empirical risk minimization, where risk is defined in terms of incentive compatible scoring rules. This approach involves mild distributional assumptions on the model, with the main requirement being a constraint on the capacity of the admissible set of choice probabilities. Further, the approach allows us to obtain bounds on the sample complexity for various models of stochastic choice i.e. the minimum number of samples needed to have a precise estimate of the true choice probabilities. This allows for distribution-free, robust estimates of choice probabilities for several well-known economic models of stochastic choice. We provide several applications and derive sample complexity upper bounds in closed form, in terms of the description and parameters of the underlying stochastic choice model.
Contact: Mary Martin at 626-395-5884 email@example.com