【摘要】 针对民调结果精度低的问题,提出了一种基于贝塔—二项概率分布模型的选举预测方法。利用历史民调数据,建立贝塔—二项概率分布模型,并采用参数估计方法,实现分布参数的估计,完成民调结果的改善。同时本模型将未表态资料考虑在内,提高了选举预测结果的精度。实验结果表明,贝塔—二项概率分布模型能够达到误差小于2%的选举预测结果。更多还原

【Abstract】 For improving the accuracy of poll result, a kind of election prediction method based on beta-binomial distribution model is proposed. By employing the poll data, the beta-binomial distribution model is established, and with the help of parameter estimation method, the distribution parameters are obtained and the poll results are improved. Then the election prediction is implemented with increased accuracy by considering unexpressed materials in the model. The experimental results show that the beta-binomial distribution model can achieve high accuracy, and the relative error is less than 2%.更多还原