國立高雄大學圖資館 |

Language: English

Back

演化式分類演算法在多核心計算環境下的設計與效率分析 = A Study ...

國立高雄大學電機工程學系碩士班

演化式分類演算法在多核心計算環境下的設計與效率分析 = A Study on the Design and Performance Analysis of a New Evolutionary Classification Algorithm in a Multi-Core Computing Environment

Record Type:	Language materials, printed : monographic
Paralel Title:	A Study on the Design and Performance Analysis of a New Evolutionary Classification Algorithm in a Multi-Core Computing Environment
Author:	洪銘宏,
Secondary Intellectual Responsibility:	國立高雄大學
Place of Publication:	[高雄市]
Published:	撰者;
Year of Publication:	民99[2010]
Description:	70面圖，表 : 30公分;
Subject:	資料探勘
Subject:	Data Mining
Online resource:	http://handle.ncl.edu.tw/11296/ndltd/50988327849357249309
Summary:	分類演算法在面對高維度大量的資料運算，通常會造成運算效率缺乏。因此本研究設計一個解析編碼之演化式分類演算法，演算法具備兩大特色，第一個特色是以解析方格的方式，解析每一個資料空間內的資料，這種方法可以處理資料點分佈複雜的問題；第二個特色為透過演化式反覆計算機制，求得較適合分類模型。為進一步提升解析編碼之演化式分類演算法的運算效率，因此本研究加入多核心程式設計方法，以多核心處理器加速解析編碼之演化式分類演算法的效率。本研究的開發了一般單核心程式的版本，與支援平行化計算的多核心程式版本，並比較單核心版本與多核心版本的效能，及分析各種不同平行化的機制間加速的效果。 This study presents GAREC (genetic algorithms with resolution-encoded chromosome) for classification. GAREC partitions the data space into resolution lattices which are encoded as chromosomes to be processed by genetic algorithms (GA.) A resolution lattice is a decision region of a class which is determined by GA. A lattice containing two or more data points belonging to different classes is considered as a new classification problem to be processed by GAREC iteratively. From the experimental results, GAREC is effective to the classification problems. In order to improve the performance of GAREC, this study implements GAREC on a multi-core computing environment using Open-MP. The performance of single- and multi-core versions of GAREC is evaluated and compared.

演化式分類演算法在多核心計算環境下的設計與效率分析 = A Study on the Design and Performance Analysis of a New Evolutionary Classification Algorithm in a Multi-Core Computing Environment
洪, 銘宏

演化式分類演算法在多核心計算環境下的設計與效率分析 = A Study on the Design and Performance Analysis of a New Evolutionary Classification Algorithm in a Multi-Core Computing Environment / 洪銘宏撰 - [高雄市] : 撰者, 民99[2010]. - 70面 ; 圖，表 ; 30公分.
參考書目：面.
資料探勘Data Mining

演化式分類演算法在多核心計算環境下的設計與效率分析 = A Study on the Design and Performance Analysis of a New Evolutionary Classification Algorithm in a Multi-Core Computing Environment
LDR:02942nam0a2200265 450 001 273035
005 20170214100141.0
009 273035
010 0 $b 精裝
100 $a 20170214y2010 k y0chiy05 e
101 1 $a chi $d chi $d eng
102 $a tw
105 $a ak am 000yy
200 1 $a 演化式分類演算法在多核心計算環境下的設計與效率分析 $d A Study on the Design and Performance Analysis of a New Evolutionary Classification Algorithm in a Multi-Core Computing Environment $f 洪銘宏撰
210 $a [高雄市] $c 撰者 $d 民99[2010]
215 0 $a 70面 $c 圖，表 $d 30公分
314 $a 指導教授：吳志宏教授
320 $a 參考書目：面
328 $a 碩士論文--國立高雄大學電機工程學系碩士班
330 $a 分類演算法在面對高維度大量的資料運算，通常會造成運算效率缺乏。因此本研究設計一個解析編碼之演化式分類演算法，演算法具備兩大特色，第一個特色是以解析方格的方式，解析每一個資料空間內的資料，這種方法可以處理資料點分佈複雜的問題；第二個特色為透過演化式反覆計算機制，求得較適合分類模型。為進一步提升解析編碼之演化式分類演算法的運算效率，因此本研究加入多核心程式設計方法，以多核心處理器加速解析編碼之演化式分類演算法的效率。本研究的開發了一般單核心程式的版本，與支援平行化計算的多核心程式版本，並比較單核心版本與多核心版本的效能，及分析各種不同平行化的機制間加速的效果。 This study presents GAREC (genetic algorithms with resolution-encoded chromosome) for classification. GAREC partitions the data space into resolution lattices which are encoded as chromosomes to be processed by genetic algorithms (GA.) A resolution lattice is a decision region of a class which is determined by GA. A lattice containing two or more data points belonging to different classes is considered as a new classification problem to be processed by GAREC iteratively. From the experimental results, GAREC is effective to the classification problems. In order to improve the performance of GAREC, this study implements GAREC on a multi-core computing environment using Open-MP. The performance of single- and multi-core versions of GAREC is evaluated and compared.
510 1 $a A Study on the Design and Performance Analysis of a New Evolutionary Classification Algorithm in a Multi-Core Computing Environment
610 0 $a 資料探勘 $a 機器學習 $a 分類演算法 $a 演化式計算 $a 平行運算 $a 多執行緖程式設計
610 1 $a Data Mining $a Machine Learning $a Classification $a Evaluable Computing $a Parallel Computing $a Multi-threading Programming
681 $a 008M/0019 $b 542201 3483 $v 2007年版
700 1 $a 洪 $b 銘宏 $4 撰 $3 483177
712 0 2 $a 國立高雄大學 $b 電機工程學系碩士班 $3 166118
801 0 $a tw $b 國立高雄大學 $c 20101227 $g CCR
856 7 $z 電子資源 $2 http $u http://handle.ncl.edu.tw/11296/ndltd/50988327849357249309