A Single-Plot Visualization of the Relative Importances of All Factors in Optimizing the Response

• 발행기관: 한국자료분석학회
• 저자명: 임성수(Sungsue Rheem)
• 간행물 정보: 『Journal of The Korean Data Analysis Society (JKDAS)』Vol.8 No.5, 1667~1673쪽, 전체 7쪽
• 주제분류: 자연과학 > 통계학
• 파일형태: PDF
• 발행일자: 2006.10.30

영문 초록

Two typical objectives of response surface methodology(RSM) are, first, to determine the optimum combination of factor levels, and second, to gain understanding of the characteristics of the response system. This article proposes a graphical display to serve that second objective of RSM. Here, the partial effect of factor j is defined as the effect of factor j when the levels of the other factors are fixed at their optimum values. The partial effects plot is a plot that shows, for j=1,...k, how the response moves as the level of factor j changes in the design region when the levels of the other factors are optimal. This plot allows graphical comparative assessment of partial effects of the factors, and so we can graphically find out which factors are relatively more important in optimizing the response and which factors are not so. A real data set from a lettuce cultivation experiment was used for illustration.

목차

1. Introduction

2. The Partial Effect Functions and Partial Effects Plot

3. Example and a Concluding Remark

References

키워드

response surface methodology graphical display partial effect function partial effects plot