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

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