论文部分内容阅读
目的 探讨流行性乙型脑炎(乙脑)发生率与预测因子的关系,为预测提供一种简单、行之有效的新方法。方法 首先计算历史上预测因子与发病率的联系数,根据联系数值的大小依次排列出最主要因子、次主要因子、再次主要因子,剔除联系数值为最小的次要因子;再把新近出现的预测因子观测值与历史上同类因子值相比较,与之最接近的因子值为该次预测用因子值,并根据该因子值与当时发病率的同一度建立预测方程,代入新因子值,解此方程得到预测值;当预测用因子数为 n′(n′≥2)时,取 n′个预测值的平均作为该次预测值。结果 应用以上预测方法预测某地某年乙脑,预测值与实际发生率很接近,仅相差0.0264/10万,准确率为97.94%。结论 可以应用基于联系数的主因子分析预测法预测乙脑。
Objective To explore the relationship between the incidence of Japanese encephalitis (JE) and predictors and to provide a simple and effective new method for prediction. The method first calculates the correlation coefficient of the historical predictor and the incidence rate, and arranges the major factor, the sub-factor, the major factor again according to the contact value, and eliminates the secondary factor with the lowest contact value; and then, the newly emerged prediction The observed value of the factor is compared with the historical value of the same kind of factor and the closest factor value is the value of the factor for this forecast, and the prediction equation is established based on the same degree of the factor and the prevalence at that time, and the new factor value is substituted into the factor value When the number of predictors is n ’(n’≥2), the average of n’ predictive values is taken as the predictive value. Results Prediction of JE in a certain year using the above prediction method was very close to the actual incidence, with a difference of 0.0264 / 100000 with an accuracy rate of 97.94%. Conclusion Prediction of Japanese encephalitis can be done by using the principal factor analysis and prediction method based on the number of links.