在日常税收统计分 析工作中，往往会对分析结果的准确性和实用性不甚满意，尤其是对分析指标进行中短期预测分析时，这问题更显得突出。在实际预测分析中使用数学模型进行线性和非线性回归测算时，多由于标准差偏大而使测算结果的可信程度降低。为了提高中短期预测时分析结果的 In the daily tax statistical analysis, the accuracy and practicability of the analysis results are often unsatisfactory, especially when analyzing the indicators in the short-term prediction and analysis. In the actual prediction analysis using mathematical models for linear and non-linear regression calculations, the more the standard deviation is too large to make the credibility of the measurement results lower. In order to improve the analysis results in the short-term forecast