Abstract:The random utility maximization model is by far the most adopted framework to estimate consumer choice behavior. However, behavioral economics has provided strong empirical evidence of irrational choice behaviors, such as halo effects, that are incompatible with this framework. Models belonging to the random utility maximization family may therefore not accurately capture such irrational behavior. Hence, more general choice models, overcoming such limitations, have been proposed. However, the flexibility of such models comes at the price of increased risk of overfitting. As such, estimating such models remains a challenge. In this work, we propose an estimation method for the recently proposed generalized stochastic preference choice model, which subsumes the family of random utility maximization models and is capable of capturing halo effects. In particular, we propose a column-generation method to gradually refine the discrete choice model based on partially ranked preference sequences. Extensive computational experiments indicate that our model, explicitly accounting for irrational preferences, can significantly boost the predictive accuracy on both synthetic and real-world data instances. Summary of Contribution: In this work, we propose an estimation method for the recently proposed generalized stochastic preference choice model, which subsumes the family of random utility maximization models and is capable of capturing halo effects. Specifically, we show how to use partially ranked preferences to efficiently model rational and irrational customer types from transaction data. Our estimation procedure is based on column generation, where relevant customer types are efficiently extracted by expanding a treelike data structure containing the customer behaviors. Furthermore, we propose a new dominance rule among customer types whose effect is to prioritize low orders of interactions among products. An extensive set of experiments assesses the predictive accuracy of the proposed approach by comparing it against rank-based methods with only rational preferences and with more general benchmarks from the literature. Our results show that accounting for irrational preferences can boost predictive accuracy by 12.5% on average when tested on a real-world data set from a large chain of grocery and drug stores.

Harmful choices

Calculated Punishment

Choosing between bad and worse: investigating choice in moral dilemmas through the lens of control

The Optimal Consumption And The Quitting Of Harmful Addictive Goods

Unbending mind: Individuals with hoarding disorder do not modify decision strategy in response to feedback under risk

Beyond pain: modeling decision-making deficits in chronic pain

Social Preferences and Deliberately Stochastic Behavior

Rationality and Psychological Accuracy of Risky Choice Models Based on Option- vs. Dimension-Wise Evaluations

Advances in understanding meso-cortico-limbic-striatal systems mediating risky reward seeking

Affective and cognitive mechanisms of risky decision making

Apparently Irrational Choice as Optimal Sequential Decision Making

Stochastic Monotonicity and Random Utility Models: The Good and The Ugly

Decision structure of risky choice

On the Estimation of Discrete Choice Models to Capture Irrational Customer Behaviors

Probability, measurement mismatches, and sacrificial moral decision-making

Perilous and unaccountable: The positive relationship between dominance and moral hazard behaviors.

Are decisions under risk malleable?

Decision-making under risk: when is utility maximization equivalent to risk minimization?

Choosing an Inferior Alternative

Optimal Consumption under a Habit-Formation Constraint: the Deterministic Case

Apparent sunk cost effect in rational agents