STOCHASTIC OPTIMIZATION APPROACH FOR PREDICTING HORSE RACING OUTCOMES

This paper uses the stochastic programming model by Peng and Uryasev (2025) to forecast horse racing outcomes. The model captures the nonlinear relationship between explanatory factors and horse running time. This random running time is characterized by quantile functions, allowing for considerable flexibility in shaping the conditional distribution as a function of factor data. The proposed model demonstrates good numerical efficiency and adaptability, making it well-suited for predicting various horse racing outcomes. The case study compares the proposed model with alternative approaches for predicting racing outcomes. The suggested model achieves comparable or superior performance while maintaining computational efficiency across various race scenarios.