凸包に基づくファジィ実時間回帰分析の構成(<一般講演>ラフ集合・分析,医療情報システムにおけるソフトコンピューティグ)

概要

論文の詳細を見る
In order to reduce the computation time in building fuzzy regression model, we propose the effective convex hull method in 1999 to decrease drastically the number of samples which are not influencial in terms of We employ a fuzzy regression analysis in building a gift-rapping method. Recently multivariant methods play a central role in data mining. In the data mining, real time data analysis is a general approach in nature. In the area, the convex hull method can provide us with a powerful tool to reduce its computational time drastically. In this paper it is explained that (1) only vertex points obtained by the convex hull are constraints in a linear programming of building a fuzzy regression model, that(2) the convex hull approach works efficiently in real-time data gathering, that (3) the number of vertexes obtained by the convex hull will not increase so much even in the real time procedure, that (4) on real time arrived data can be effectively and efficinetly treated by the convex hull method in fuzzy regression analysis, that (5) the method does not consume the duplicated computation of the past analysis, and that (6) the analysis can be built on the basis of the past one using newly arrival data.
2001-10-27