研究对称熵损失下成功概率P的Bayes估计和E—Bayes估计,证明了前者的存在性及唯一性.模拟结果表明E-Bayes估计优于极大似然估计和Bayes估计.并将E—Bayes方法应用在证券投资预测之中,预测效果较好.
The authors studied the Bayesian and E-Bayesian estimation of probability of success p under symmetric entropy loss function and proved the existence and uniqueness of the Bayesian estimation. The simulation results showed that the E-Bayesian estimator outperformed than the maximum likelihood estimator and Bayesian estimator. The E-Bayesian method was applied to the forecast of the stock investment and the results of forecasting were precise.