Robust Bayes Forecasting For Grouped Binary Data With Known Distortion Interval

Abstract

The paper is devoted the problem of robust forecasting for the beta-mixed hierarchical models of grouped binary data in the case of stochastic additive distortions of binary observations. In the case of known lower and upper bounds of the distortion intervals, a new robust minimax Bayes predictor is developed. The performance of the proposed forecasting technique is validated by computer simulationtadata.