By use of the super-club data and entropy of the information table, the discretization of continuous attributes was implemented. 根據條件屬性與決策屬性之間的依賴(lài)關(guān)系及相對信息熵概念，實(shí)現了連續屬性的離散化。
Since conditional attribute set has been an algebraic consistent set, it can guarantee to obtain a reduction by top-bottom strategy to eliminate reducible attributes. 由于所有條件屬性集已是一個(gè)代數協(xié)調集，采用自頂而下逐步刪除可約屬性的策略，提出一種基于數據庫系統求代數約簡(jiǎn)的算法。
