國立高雄大學圖資館 |

語系: 繁體中文

說明(常見問題)

圖資館首頁

登入

回首頁

切換: 標籤 | MARC模式 | ISBD

應用多重支持之廣義關聯分類法建構大學休退學預測系統 = Applying...

國立高雄大學資訊工程學系碩士班

應用多重支持之廣義關聯分類法建構大學休退學預測系統 = Applying Multi-Supported Generalized Associative Classification to Constructing University Dropout Prediction System

紀錄類型:	書目-語言資料,印刷品 : 單行本
並列題名:	Applying Multi-Supported Generalized Associative Classification to Constructing University Dropout Prediction System
作者:	鄭光盛,
其他團體作者:	國立高雄大學
出版地:	[高雄市]
出版者:	撰者;
出版年:	2013[民102]
面頁冊數:	60面圖，表 : 30公分;
標題:	廣義關聯分類規則
標題:	Generalized Associative Classification
電子資源:	http://handle.ncl.edu.tw/11296/ndltd/07079437330698011170
附註:	參考書目：面57-60
附註:	103年12月16日公開
摘要註:	學生休退學預測問題是教育資料探勘研究領域中很受重視的問題，因為社會結構的轉變，台灣的出生人口數逐年減少。再加上一般大眾收入的減少、工時增長、物資價格的調升，以及高學歷畢業生失業率逐年攀升，使得民眾不再信任高學歷是獲得工作的保證。進而造成大專院校招生不足的現象愈趨嚴重。為維持一定的學生人數與學校辦學的品質，如何事先查覺學習不太佳的學生，對可能會休學或退學的學生進行早期的輔導與補救措施，就成為各大學普遍關注的問題。因為教育的多元化，學生休退學不再如過去只是單純的成績或是經濟問題所造成，大多數都是零散但具有關聯性的成因。過去以人工一一來判讀這些原因，曠日廢時且不具成效。本研究的主要目的即在發展一個適用於大學生休退學的預測系統，本系統預測方法採用我們設計的多重支持之廣義關聯分類法(GACMS)。我們的GACMS方法，以關聯分類法中的 CMAR方法為基礎，輔以多重支持度機制讓支持度不高卻很重要的規則可以被系統學習到，再加上納入資料項目間存在的階層式資訊，讓適用性較窄的規則有機會彙總為強度夠的廣泛規則。經由實驗結果，GACMS方法可以得到比現行採用的預測方法更高的精準度。且能找出重要但容易被其他方法忽略的規則。 Since the birth rate decreasing is changing the civilian age structure in Taiwan, the student dropout prediction has become an important issue in the educational data mining research community. The income cutback, long working hours, inflation and unemployment rate waiting in the society for university graduate and under graduate students are making students no longer trust that the university diplomas are guarantee for better working opportunities, and it makes the universities not easy to enroll enough new students. In order to keep the student number and education quality to a competitive level, all of the universities are concerning about how to discover the students with learning problems and then engage consolation and compensation works in the early stage.The reasons for student dropout are no longer caused only by academic grades or personal financial problems, since the modern university education is taking routes in various directions. The reasons are diverse yet still event related. In the past, they are examined case by case through manpower that is time consuming and ineffective. This research is to develop a system suitable for university dropout prediction by adopting our Generalized Associative Classification with Multiple Minimum Supports (GACMS).Our GACMS method is based on the CMAR method, a well known associative classification method. With the help of multi-supported method, the system can learn less-supported but important rules. Adding the taxonomy information existed in the data items, the less adoptable rules could become dominating common rules.According to the experimental results, GACMS is capable of acquiring better precision in dropout prediction compared to using the traditional prediction methods, and it also finds important rules that are usually to be ignored by other methods.

應用多重支持之廣義關聯分類法建構大學休退學預測系統 = Applying Multi-Supported Generalized Associative Classification to Constructing University Dropout Prediction System
鄭, 光盛

應用多重支持之廣義關聯分類法建構大學休退學預測系統 = Applying Multi-Supported Generalized Associative Classification to Constructing University Dropout Prediction System / 鄭光盛撰 - [高雄市] : 撰者, 2013[民102]. - 60面 ; 圖，表 ; 30公分.
參考書目：面57-60103年12月16日公開.
廣義關聯分類規則Generalized Associative Classification

應用多重支持之廣義關聯分類法建構大學休退學預測系統 = Applying Multi-Supported Generalized Associative Classification to Constructing University Dropout Prediction System
LDR:04678nam0a2200289 450 001 389691
005 20170214100943.0
009 389691
010 0 $b 精裝
010 0 $b 平裝
100 $a 20170214d2013 k y0chiy05 e
101 1 $a chi $d chi $d eng
102 $a tw
105 $a ak am 000yy
200 1 $a 應用多重支持之廣義關聯分類法建構大學休退學預測系統 $d Applying Multi-Supported Generalized Associative Classification to Constructing University Dropout Prediction System $z eng $f 鄭光盛撰
210 $a [高雄市] $c 撰者 $d 2013[民102]
215 0 $a 60面 $c 圖，表 $d 30公分
300 $a 參考書目：面57-60
300 $a 103年12月16日公開
314 $a 指導教授：林文揚博士
328 $a 碩士論文--國立高雄大學資訊工程學系碩士班
330 $a 學生休退學預測問題是教育資料探勘研究領域中很受重視的問題，因為社會結構的轉變，台灣的出生人口數逐年減少。再加上一般大眾收入的減少、工時增長、物資價格的調升，以及高學歷畢業生失業率逐年攀升，使得民眾不再信任高學歷是獲得工作的保證。進而造成大專院校招生不足的現象愈趨嚴重。為維持一定的學生人數與學校辦學的品質，如何事先查覺學習不太佳的學生，對可能會休學或退學的學生進行早期的輔導與補救措施，就成為各大學普遍關注的問題。因為教育的多元化，學生休退學不再如過去只是單純的成績或是經濟問題所造成，大多數都是零散但具有關聯性的成因。過去以人工一一來判讀這些原因，曠日廢時且不具成效。本研究的主要目的即在發展一個適用於大學生休退學的預測系統，本系統預測方法採用我們設計的多重支持之廣義關聯分類法(GACMS)。我們的GACMS方法，以關聯分類法中的 CMAR方法為基礎，輔以多重支持度機制讓支持度不高卻很重要的規則可以被系統學習到，再加上納入資料項目間存在的階層式資訊，讓適用性較窄的規則有機會彙總為強度夠的廣泛規則。經由實驗結果，GACMS方法可以得到比現行採用的預測方法更高的精準度。且能找出重要但容易被其他方法忽略的規則。 Since the birth rate decreasing is changing the civilian age structure in Taiwan, the student dropout prediction has become an important issue in the educational data mining research community. The income cutback, long working hours, inflation and unemployment rate waiting in the society for university graduate and under graduate students are making students no longer trust that the university diplomas are guarantee for better working opportunities, and it makes the universities not easy to enroll enough new students. In order to keep the student number and education quality to a competitive level, all of the universities are concerning about how to discover the students with learning problems and then engage consolation and compensation works in the early stage.The reasons for student dropout are no longer caused only by academic grades or personal financial problems, since the modern university education is taking routes in various directions. The reasons are diverse yet still event related. In the past, they are examined case by case through manpower that is time consuming and ineffective. This research is to develop a system suitable for university dropout prediction by adopting our Generalized Associative Classification with Multiple Minimum Supports (GACMS).Our GACMS method is based on the CMAR method, a well known associative classification method. With the help of multi-supported method, the system can learn less-supported but important rules. Adding the taxonomy information existed in the data items, the less adoptable rules could become dominating common rules.According to the experimental results, GACMS is capable of acquiring better precision in dropout prediction compared to using the traditional prediction methods, and it also finds important rules that are usually to be ignored by other methods.
510 1 $a Applying Multi-Supported Generalized Associative Classification to Constructing University Dropout Prediction System $z eng
610 0 $a 廣義關聯分類規則 $a 多重支持度 $a 階層架構關聯規則 $a 學生休退學預測 $a 關聯分類法
610 1 $a Generalized Associative Classification $a Multiple Minimum Supports $a Hierarchy $a Student Dropout Prediction $a CMAR
681 $a 008M/0019 $b 464103 8795 $v 2007年版
700 1 $a 鄭 $b 光盛 $4 撰 $3 614551
712 0 2 $a 國立高雄大學 $b 資訊工程學系碩士班 $3 353878
801 0 $a tw $b NUK $c 20141204 $g CCR
856 7 $z 電子資源 $2 http $u http://handle.ncl.edu.tw/11296/ndltd/07079437330698011170