The Fractional Factorial Design in Latin Sqauare Designs and in Crossover Designs
碩士 === 高雄師範大學 === 數學研究所 === 99 === We arrange the treatment combinations of the 2k−l design into a Graeco-Latin square and estimate the corresponding effects. The idea is useful for the large factors in an experiment. Actually, the fractional factorial design associated with the Graeco-Latin square m...
Main Authors: | , |
---|---|
Other Authors: | |
Format: | Others |
Language: | en_US |
Published: |
2011
|
Online Access: | http://ndltd.ncl.edu.tw/handle/95415159895711451071 |
id |
ndltd-TW-099NKNU5479021 |
---|---|
record_format |
oai_dc |
spelling |
ndltd-TW-099NKNU54790212015-10-13T20:13:38Z http://ndltd.ncl.edu.tw/handle/95415159895711451071 The Fractional Factorial Design in Latin Sqauare Designs and in Crossover Designs 部分因子設計在拉丁矩陣與交叉設計上 Yuan-Chih Lu 盧淵智 碩士 高雄師範大學 數學研究所 99 We arrange the treatment combinations of the 2k−l design into a Graeco-Latin square and estimate the corresponding effects. The idea is useful for the large factors in an experiment. Actually, the fractional factorial design associated with the Graeco-Latin square may decrease the number of observations. That is, the design is more economical. After introducing the Graeco-Latin square, we show the Crossover design. In this article, we use the fold-over technique to construct a crossover design with two rows, and then arrange the treatment combinations into these replicated 2 × 2 Latin squares properly. The designs not only estimate the effects but also reduce the influence of residual effect. Finally, we make a regression analysis and display the factor effects by the corresponding regression coefficients. Pi-Hsiang Huang 黃必祥教授 2011 學位論文 ; thesis 34 en_US |
collection |
NDLTD |
language |
en_US |
format |
Others
|
sources |
NDLTD |
description |
碩士 === 高雄師範大學 === 數學研究所 === 99 === We arrange the treatment combinations of the 2k−l design into a Graeco-Latin square and estimate the corresponding effects. The idea is useful for the large factors in an experiment. Actually, the fractional factorial design associated with the Graeco-Latin square may decrease the number of observations. That is, the design is more economical. After introducing the Graeco-Latin square, we show the Crossover design. In this article, we use the fold-over technique to construct a crossover design with two rows, and then arrange the treatment combinations into these replicated 2 × 2 Latin squares properly. The designs not only estimate the effects but also reduce the influence of residual effect. Finally, we make a regression analysis and display the factor effects by the corresponding regression coefficients.
|
author2 |
Pi-Hsiang Huang |
author_facet |
Pi-Hsiang Huang Yuan-Chih Lu 盧淵智 |
author |
Yuan-Chih Lu 盧淵智 |
spellingShingle |
Yuan-Chih Lu 盧淵智 The Fractional Factorial Design in Latin Sqauare Designs and in Crossover Designs |
author_sort |
Yuan-Chih Lu |
title |
The Fractional Factorial Design in Latin Sqauare Designs and in Crossover Designs |
title_short |
The Fractional Factorial Design in Latin Sqauare Designs and in Crossover Designs |
title_full |
The Fractional Factorial Design in Latin Sqauare Designs and in Crossover Designs |
title_fullStr |
The Fractional Factorial Design in Latin Sqauare Designs and in Crossover Designs |
title_full_unstemmed |
The Fractional Factorial Design in Latin Sqauare Designs and in Crossover Designs |
title_sort |
fractional factorial design in latin sqauare designs and in crossover designs |
publishDate |
2011 |
url |
http://ndltd.ncl.edu.tw/handle/95415159895711451071 |
work_keys_str_mv |
AT yuanchihlu thefractionalfactorialdesigninlatinsqauaredesignsandincrossoverdesigns AT lúyuānzhì thefractionalfactorialdesigninlatinsqauaredesignsandincrossoverdesigns AT yuanchihlu bùfēnyīnzishèjìzàilādīngjǔzhènyǔjiāochāshèjìshàng AT lúyuānzhì bùfēnyīnzishèjìzàilādīngjǔzhènyǔjiāochāshèjìshàng AT yuanchihlu fractionalfactorialdesigninlatinsqauaredesignsandincrossoverdesigns AT lúyuānzhì fractionalfactorialdesigninlatinsqauaredesignsandincrossoverdesigns |
_version_ |
1718045207425974272 |