建立预测模型.ppt

Chapter 3Introduction to Predictive Modeling : From Correlation to Supervised Segmentation 建立預測模型 --- 從關連到監督式區隔;;;In the process of discussing supervised segmentation, we introduce one of the fundamental ideas of data mining: finding or selecting important, informative variables or “attributes” of the entities described by the data. 在我們討論監督式區隔時，我們講到一個資料探勘的基本觀念---由資料中找出重要且具傳遞訊息能力的變數或屬性 What exactly it means to be “informative” varies among applications, but generally, information is a quantity that reduces uncertainty about something. 不同的應用，「具傳遞訊息能力」可能所指不同，但是「資訊」本身就是能對某項事情減少不確定性 ;Finding informative attributes also is the basis for a widely used predictive modeling technique called tree induction, which we will introduce toward the end of this chapter as an application of this fundamental concept. 找出具傳遞訊息能力屬性是一個廣泛使用的建立預測模型方法的基礎，這方法叫做決策樹歸納法，本章將慢慢介紹。 ;;Outline;Models, Induction, and Prediction;Models, Induction, and Prediction;Models, Induction, and Prediction;Models, Induction, and Prediction; Terminology : Induction and deduction;Outline;Supervised Segmentation 監督式區隔;Supervised Segmentation監督式區隔;Supervised Segmentation監督式區隔;Outline;Selecting Informative Attributes選擇具傳遞訊息能力的屬性 ;Selecting Informative Attributes選擇具傳遞訊息能力的屬性 ;Selecting Informative Attributes ;Selecting Informative Attributes ;Selecting Informative Attributes ;Selecting Informative Attributes ;;Selecting Informative Attributes ;Selecting Informative Attributes ;Selecting Informative Attributes ;Selecting Informative Attributes ;Selecting Informative Attributes ;Selecting Informative Attributes ;Selecting Informative Attributes ;Selecting Informative Attributes ; Numeric variables;Outline;Example: Attribute Selection with Information Gain用Information Gain去選取屬性;Example: Attribute Selection with Information Gain 用Information Gain去選取屬性;香菇資料集;Example: Attribute Selection with Information Gain 用Information Gain去選取屬性;Example: Attribute Selection with Information Gain 用Information Gain去???取