Compilation method of two class taxonomic forecasting models, compilation device of program and two class taxonomic forecasting models for classification forecasting model compilation

摘要

With simple operation, substantially doing be close to two class classifications 100%, it draws up two class taxonomic forecasting models whose reliability is high. Two class taxonomic forecasting models to calculate the discriminant function in order to classify into two classes which decide study sample set beforehand on the basis of the explanatory variable which it occurs a) concerning each sample of study sample set, to calculate the distinction score of each study sample b) on the basis of the discriminant function which was calculated, to decide the correction of classification estimate of each study sample c) on the basis of the distinction score which was calculated, to decide the error taxonomic sample territory d) error maximum it is in the midst of study sample setting which is classified, on the basis of the smallest distinction score, e) removing the study sample which is included in the error taxonomic sample territory, newIt forms study sample set, it is drawn up by repeating f) vis-a-vis new study sample set, a) - e).