By means of the law of large number , repeated sampling method and nonlinear regression method , the direct component models are built up for bronze inscription system , seal script system , official script system and song dynasty regular script system 借助于大數(shù)定律、重復抽樣方法和非線性回歸方法,分別建立了金文、小篆、漢隸和宋楷幾個具有代表性的、不同時代的漢字系統(tǒng)直接構件關系模型。