• Journal of the Korean Statistical Society
• /
• 제11권1호
• /
• pp.1-11
• /
• 1982

Srivastava and Anderson(1970) illustrate a method of obtaining Balanced (but not orthogonal) Resolution $IV^*$ designs starting with a BIB design. The incidence matrix of a BIB design with parameters (v, b, r, k, and $\lambda$) is utilized to obtain Balanced Resolution $IV^*$ designs with m factors and n=2b runs, where $m \leq v$. In this paper, the same idea is extended to the case of PBIB designs to obtain Partially Balanced Resolution $IV^*$ designs. In the designs obtained here the variances are balanced and the covariances are partially balanced with respect to the main effects. A proof of this property of Partially Balanced Resoultion $IV^*$ designs is given. The efficiency of Partially Balanced Resolution $IV^*$ designs is also considered and examples of Partially Balanced Resoultion $IV^*$ designs are included.

• 대한안전경영과학회지
• /
• 제7권5호
• /
• pp.69-84
• /
• 2005

This paper proposes statistically designed experiments which provide a proactive means to implement safety and environmental applications. Minimal experimental designs such as fractional factorial design, Plackett-Burman design, Box-Behnken design are economical and can be achieved tremendous savings with relatively few experiments. These experimental designs and analysis methods are illustrated with cases.

• Journal of the Korean Statistical Society
• /
• 제34권2호
• /
• pp.141-151
• /
• 2005

Under the assumption that the three-level factors are quantitative, the linear effects are taken more attention than the quadratic effects of the interaction terms. Webb (1971) presented some small incomplete factorial designs that are mixed two- and three-level designs with 20 or fewer runs. The designs provided the estimating linear-by-linear components of interactions between the three-level factors; moreover, they could also offer estimation of interactions that interest the experiments. Webb used ad hoc methods to find these plans; hence, there was still no unified structure to those experiments. In this paper, we develop the methods to construct the $2^{n}3^3$ and $2^{1}3^3$ designs. The designs constructed by these methods not only supply orthogonal estimates of all the main effects but also permit estimation of all the two-factor interactions not involving the quadratic effects. Furthermore, the designs we find are nearly orthogonal.

• Journal of Korean Society for Atmospheric Environment
• /
• 제24권E2호
• /
• pp.83-91
• /
• 2008

This work investigated the methodology to simultaneously optimize the ozone and relative humidity composition for the $NO_x$ degradation using soil biofilter. Experiments were made as a function of inlet ozone concentration ($0{\sim}1,770\;ppb$) and relative humidity ($38{\sim}81%$). Factorial design ($2^2+3$) and response surface methodology by central composite designs were used to examine the role of two factors and optimal response condition on $NO_x$ degradation. It was found that a second-order response surface model can properly interpret the experimental data with an $R^2$-value of 0.9730 and F-value of 71.83, based on which the maximum $NO_x$ degradation was predicted up to 92.8% within our experimental conditions.

• 한국데이터정보과학회:학술대회논문집
• /
• 한국데이터정보과학회 2006년도 PROCEEDINGS OF JOINT CONFERENCEOF KDISS AND KDAS
• /
• pp.121-129
• /
• 2006

Two color or cDNA microarrays are extensively used to study relative expression levels of thousands of genes simultaneously. 0かy two tissue samples can be hybridized on a single microarray slide. Thus, a microarray slide necessarily forms an incomplete block design with block size two when more than two tissue samples are under study. We also need to control for variability in gene expression values due to the two dyes. Thus, red and green dyes form the second blocking factor in addition to slides. General design problem for these microarray experiments is discussed in this paper. Designs for factorial cDNA microarrays are also discussed.

• Journal of the Korean Statistical Society
• /
• 제3권2호
• /
• pp.85-101
• /
• 1974

In a Balanced Factorial Experiments (BFE) with n factors $F_1, F_2,\cdots,F_n$ at $m_1, m_2,\cdots,m_n$ levels respectively, Shah [15] has considered the following association scheme: the two treatments are the $(P_1, P_2,\cdot,P_n)$th associates, where $p_i=1$ if the ith factor occurs at the same level in both treatments and $p_i=0$ otherwise; $\lambda_{(p_1,p_2,\cdots,p_n)}$ will denote the number of times these treatments occur together in a block. He has showed that a BFE is partially Blanced Incomplete Block(PBIB) design with repsect to the above association scheme. Kurjian and Zelan [6] have proved that factorial designs possessing a Property A (a particular structure for their matrix NN') are factorially balanced.

• Journal of the Korean Data and Information Science Society
• /
• 제17권4호
• /
• pp.1053-1065
• /
• 2006

This paper proposes a method in one-way layouts for longitudinal data with informative drop-out. When dropouts are informative, that is, correlated with unobserved data and/or the previous observed data, the simple imputation methods such as 'last observation carried forward' (LOCF) methods would arise the bias of the testing models. The maximum likelihood procedure combined with a logit model for the drop-out process is proposed to test treatment effects for one factorial designs and compared with LOCF method in two examples.

• 대한가정학회지
• /
• 제23권3호
• /
• pp.49-53
• /
• 1985

The typical texture profile analysis of acorn flour gel was investigated with Instron univ. testing machine by two experimental designs, 3$\times$3 latin square and $3^{3}$factorial experiment. As the result, it was revealed that Latin square is a useful method to reduce the number of experiments, in the case of a negligible interaction.

• 산업식품공학
• /
• 제15권2호
• /
• pp.97-115
• /
• 2011

이 총설에서는 반응표면분석법을 이용하여 식품제조프로세스를 최적화하는 방법에 대하여 검토하였다. 반응표면분석을 수행하기 위한 절차와 반응표면분석의 필수적인 기본이론을 설명하였고, 반응표면분석법 중에서도 대부분 사용되는 2차 실험계획법(3인자 완전요인, 중심합성, Box-Behnken, 및 Doehlert 계획법)들에 대한 장단점 및 효율성을 비교하였다. 식품제조프로세스를 최적화하는데 반응표면분석법을 적용하기 위해서는 우선 실험계획을 선택하고, 적절한 모델함수를 적합화한 다음, 적합화된 모델의 질 및 실험데이터와의 예측의 정확성을 평가할 필요가 있다. 식품제조프로세스를 최적화할 때 일부요인계획, 완전요인계획 및 Plackett-Burman 계획 등과 같은 실험계획법을 사용하여 중요한 실험인자를 일차적으로 스크린한 다음, 2차 실험계획법을 선택하는 것이 바람직하다. 실제적으로 최적실험조건을 찾기 위해서는 F-test, 수정 $R^2$ 등과 같은 분산분석을 사용하여 모델을 적합화하는 것이 바람직하다. Doehlert 계획과 Box-Behnken 계획은 중심합성계획법보다 좀 더 효율적이며, 최근에는 이 계획들을 적용한 문헌의 수가 증가하고 있는 추세이다. 더욱이 이 계획들은 3수준 완전요인계획법보다는 비교할 필요도 없이 훨씬 더 효율적이다. Box-Behnken설계는 식품분야에서와 같이 극한조건(즉, 인자들이 동시에 가장 높거나 혹은 가장 낮은 수준의 실험 조건)하에서 실험을 하는 것을 피하고자 할 때 유용하다. Doehlert 계획에서는 각 인자들의 수준(level)이 다르기 때문에, 몇몇 인자들이 가격적인 면에서 그리고(혹은) 장비사용에 제약을 받는 제한이 있다든지 혹은 인자의 중요도에 따라 수준의 수를 조절해야 할 필요가 있을 때에는 Doehlert 계획이 아주 유용하다. 종래에는 반응표면분석법의 2차 회귀모델 실험계획법 중에서 다른 계획법(Box-Behnken 계획 및 Doehlert 계획)에 비해 중심합성계획법을 압도적으로 많이 적용해 왔다. 그러나 Box-Behnken 계획 및 Doehlert 계획은 중심합성계획법보다 장점이 많기 때문에, 향후에는 Doehlert 계획과 Box-Behnken 계획을 사용하여 식품제조프로세스를 최적화하는 쪽으로 초점이 맞추어 지리라고 전망한다.

• Journal of the Korean Statistical Society
• /
• 제31권3호
• /
• pp.277-299
• /
• 2002

Most statistical designs, such as orthogonal designs and optimal designs, are based on a specific statistical model. It is very often that the experimenter does not completely know the underlying model between the response and the factors. In computer experiments, the underlying model is known, but too complicated. In this case we can treat the model as a black box, or model to be unknown. Both cases need a space filling design. The uniform design is one of space filling designs and seeks experimental points to be uniformly scattered on the domain. The uniform design can be used for computer experiments and also for industrial experiments when the underlying model is unknown. In this paper we shall introduce the theory and method of the uniform design and related data analysis and modelling methods. Applications of the uniform design to industry and other areas are discussed.