A connection number-based principal factor analysis forecast method to forecast the encephalitis B epidemics / 中华流行病学杂志

ABSTRACT

<p><b>OBJECTIVE</b>To detect the relations between incidence rate of the epidemical encephalitis B and related factors, to provide a simple, valid and practical new method for forecasting encephalitis B eipdemics.</p><p><b>METHODS</b>Connection number between the incidence rate of encephalitis B and the historical forecast factors was computed, before ranking the first, second and the third principal factor, to remove the factor with the smallest value in the light of the connection number before comparing the newest value of forecast factors with the same kind of history while the most nearly value becoming the forecasting factor value and to establish a forecasting equation according to the factor value and the consistent degree of the incidence rate of encephalitis B at that time. Finally, to put into the new factor value to get this forecast value under this equation. Assuming that there are n' (n' >or= 2) forecast factors, this time forecast value can then be directly obtained from the average of these estimate values.</p><p><b>RESULTS</b>Using above forecast method to forecast the incidence rate of encephalitis B at certain place and year, the predicting value is very much close to the actual incidence rate. Difference between the predicting value forecasted by the above-mentioned method and the actual incidence rate is only 0.0264/100 000 with an accurate rate of 97.94%.</p><p><b>CONCLUSION</b>This principal factor analysis forecast method based on connection number in forecasting the incidence rate of encephalitis B prevention is acceptable.</p>