tailieunhanh - Xây dựng cây quyết định sử dụng phụ thuộc hàm xấp xỉ

In this paper, we introduce in brief on the concepts of Approximate Functional Dependency (AFD), of Approximate Functionally Cross-Characteristic Dependency known as type II AFD and describe an adoption of AFD to a constructing method of decisson tree for databases mining purposes. | ’ Tap ch´ Tin hoc v` Diˆu khiˆn hoc, , (2007), 179–186 ı e e . . a ` . . ´ ’. ´ ’ ˆ ˆ ˆ ˆ ` ˆ XAY DU NG CAY QUYET DINH SU DUNG PHU THUOC HAM XAP XI . . . . . ˆ ˆ ˜ ´. A VU DU C THI, TR` N QUANG DIEU . Viˆn Cˆng nghˆ thˆng tin, Viˆn Khoa hoc v` Cˆng nghˆ Viˆt Nam e o e o e e e . . . . a o . . Abstract. In this paper, we introduce in brief on the concepts of Approximate Functional Dependency (AFD), of Approximate Functionally Cross-Characteristic Dependency known as type II AFD and describe an adoption of AFD to a constructing method of decisson tree for databases mining purposes. ´ ´ e a e o a a ’ e T´m t˘t. Trong b`i b´o n`y, ch´ ng tˆi gi´.i thiˆu so. vˆ kh´i niˆm phu thuˆc h`m xˆ p xı, phu o a a a a u o o . ` . . . . . ´ ´ ´ a o u a o ınh ’ o e o e thuˆc h`m xˆ p xı liˆn quan dˆn quan h`m sˆ gi˜.a c´c thuˆc t´ cua mˆt quan hˆ (Phu thuˆc o a a ’ e . . . . . . ´ ´ ´ h`m xˆ p xı loai hai) v` dung phu thuˆc h`m xˆ p xı nh˘ m xˆy cˆy quyˆt dinh trong khai a a ’ . a´ a e . o a a ’ ` a a . . . . ph´ d˜. liˆu. a u e . ˆ ´. 1. GIO I THIEU . ´ a a Phu thuˆc h`m xˆ p xı (Approximate Functional Dependency AFD) v` ph´p ph´t o a a ’ a . . . nhiˆu t´c gia dˆ cˆp v` dung trong nhiˆu b`i ` a ` ´ ’ ` a a´ e e a e . hiˆn c´c phu thuˆc h`m xˆ p xı d˜ du . e a o a a ’ a . . . . . liˆu ([1,3]). Phu thuˆc h`m xˆ p xı l` mˆt phu thuˆc h`m c´ t´ chˆ t gˆn ´ ’ a o ´ a to´n khai ph´ d˜ e a a u . o a a o a o ınh a ` . . . . . .i mˆt quan hˆ r v` dinh ngh˜ nhu. sau. ´ d´ng dˆi v´ o u o o ıa e a . . . . ´ Dinh ngh˜ 1. Phu thuˆc h`m xˆ p xı (Approximate Functional Dependency - AFD) ıa o a a ’ . . . ´ ’ e ´ Cho ε, 0 ε 1, X → Y l` phu thuˆc h`m xˆ p xı nˆu: approx(X → Y ) a o a a ε, v´.i o . . ’ a r v` X → Y d´ng trˆn s}/|r|). u e approx(X → Y ) = 1 (max {|s|, s l` tˆp con cu a a a . ’ . dˆy, |s|, |r| l` sˆ phˆn tu. cua s v` r. ´ ` ’ ’ a a o a O a ´ ’ y . e ’ Trong c´c b`i to´n quan l´ tˆ, thu.`.ng xay ra c´c nh´m thuˆc t´ c´ su.