量子社会システム分析法の確率力学的基礎付け(一般講演,医療の将来とソフトコンピューティング)

抄録

A new model is proposed to analyze social system from their macroscopic point of view. In this method, probability distribution is requested to satisfy the quadratic derivative equation of Schrondinger type. We thought that the data value mean the fuzzy number. In order to treat properly the ambiguity of data, we introduce fuzzy derivative and fuzzy method. The quadratic derivative of the continuous data do not change largely. So, the average of the quadratic derivative is considered good for interpolation of quadratic derivative. The fuzzy integration is used to predict the ambiguity range of the predicted values. Several examples shows that the method works well to give a smoothed curve going to through the data points. The method is also applied to prediction of medical disease in Japan. Taking the age distributions of three disease as examples, we show that the real potential represent the structural characteristics of stationary distribution and imaginary potential indicate the time-change of the absolute value and the shape of distribution. Interactions between distribution are also introduce and it is shown that its approximate solution include a simple relation of net transfer. An idea of the politics is proposed as tool of system analysis and useless is shown. It is merit that we can recognize easily the characteristic structure or dynamics of system by pattern