I am a Ph.D. candidate in Economics at the University of California, Irvine (UCI). I am on the 2025-26 job market. I specialize in decision theory, experimental economics, and behavioral economics. My current research focuses on devising efficient algorithms to fit utility functions to choice data from the real world and experiments in the lab.

with Igor Kopylov

Abstract: We model stochastic choice rules via finitely many types θ that maximize distinct expected utility functions  and use  endogenous tie-breaking rules. First, we characterize discrete random expected utility (DREU) where the likelihood μ(θ) of each relevant type θ is preserved across all menus A. This model is a discrete version for the random expected utility of  Gul and Pesendorfer (2006), but our axioms, identification, and tie-breaking procedures are novel. More generally, we propose discrete-map expected utility (DMEU), where the likelihoods μA(θ) are contingent on the menu A. The continuous map μA is identified uniquely in our model. DMEU captures various kinds of context dependence, such as reason-based choice, extremeness aversion, and other behavioral patterns. Moreover, we use  DMEU to model self-selection, where types can increase their participation rates across distinct menus,  but only if  their best choices are improved. The standard monotonicity principle for stochastic choices delivers a novel self-selective property for the type likelihoods  μA. This bias is identifiable because all types are assumed to maximize expected utility functions. By contrast, it can distort the likelihoods μ(θ) without turning Block-Marschak polynomials negative in the general random utility model. Finally, we discuss applications to random risk aversion and random  Cobb-Douglas utility.