Synergistic Genetic Algorithms for Inductive Learning (SynGALL)

Name: รัตนา จรัสศุภวัฒน์

Organization : มหาวิทยาลัยเทคโนโลยีพระจอมเกล้าธนบุรี

keyword: Classification

; GAIL

; Genetic Algorithms

; Genetic Algorithms for Inductive Learning

; Inductive Learning

; Information Technology

; King Mongkut's University of Technology Thonburi

; KMUTT

; Synergistic Genetic Algorithms for Inductive Learning

; Synergy

; SynGAIL

Abstract: Genetic Algorithms for Inductive Learning (GAIL) is a technique in Data Mining which can be used for classification. It combines genetic algorithms with inductive learning to produce rules which can efficiently classify data set into appropriate categories.This research study applied and improved GAIL in classification of students' data. Synergy of different GAILs was implemented which resulted in Synergistic Genetic Algorithms for Inductive Learning (SynGAIL). A new fitness function was also introduced to accord withthe characteristics of the data. SynGAIL was applied to obtain rules which can be used to estimate the ability to study (i.e. the range of GPA expected) of applicants to the M.Sc. programme in Information Technology at School of Information Technology, King Mongkut'sUniversity of Technology Thonburi. Data set used for training and testing were records of students who had completed the programme between semester 2/1996 and semester 1/1999.Among results obtained from GAIL, SynGAIL and SynGAIL (with new fitness function), the third yielded the best result (with an accuracy of 74%). The study reveals relevant characteristics in the classification and highlights the importance of having adequate amount of training data set. The study also affirms the potential of synergistic approach, especially in improving efficiency of the result when training is restricted by small amount of data set.

มหาวิทยาลัยเทคโนโลยีพระจอมเกล้าธนบุรี

Address: กรุงเทพมหานคร (Bangkok)

Email: info.lib@mail.kmutt.ac.th

Name: ดร. กิตติชัย ลวันยานนท์

Role: อาจารย์ที่ปรึกษา

Created: 2543

Issued: 2005-10-11

Modified: 2006-12-25

วิทยานิพนธ์/Thesis

ISBN: 9746249185

CallNumber: INT483

tha

eng

DegreeName: วิทยาศาสตรมหาบัณฑิต

Level: ปริญญาโท

Descipline: เทคโนโลยีสารสนเทศ

Grantor: มหาวิทยาลัยเทคโนโลยีพระจอมเกล้าธนบุรี

©copyrights King Mongkut's University of Technology Thonburi

ลำดับที่.	ชื่อแฟ้มข้อมูล	ขนาดแฟ้มข้อมูล	จำนวนเข้าถึง	วัน-เวลาเข้าถึงล่าสุด
1	INT483.pdf	20.09 MB	138	2024-09-12 13:50:30
2	INT483ab.pdf	97.04 KB	20	2020-06-05 12:02:14
3	INT483ab.txt	1.87 KB	21	2020-06-05 12:02:28

ใช้เวลา

0.025773 วินาที