... functions(Hanson and Burr [HB88]), dynamic adjustment of the network topology (Me´zard and Nadal [MN89], Fahlman and Lebiere [FL90], Le Cun, Denker, and Solla [LDS90], and Harp, Samad, and Guha [HSG90] ), and ... data in preparation for classification and prediction can involve data cleaning to reduce noise or handle missing values, relevance analysis to removeirrelevant or redundant attributes, and data ... the data. Early decision tree algorithms typi-cally assume that the data are memory resident—a limitation to data mining on largedatabases. Several scalable algorithms, such as SLIQ, SPRINT, and...