본문 바로가기 주메뉴 바로가기
국회도서관 홈으로 정보검색 소장정보 검색

결과 내 검색

동의어 포함

목차보기

Title Page

Abstract

Contents

Nomenclature 11

CHAPTERS I. Introduction 15

A. Introduction 15

1. Process design 15

2. Experimental design 16

a. Mixture experiments for pharmaceutical formulation 17

3. Robust design 18

4. Pharmaceutical process 21

B. Problem statement, research motivation, and significance 22

1. Problem statement 22

2. Research motivation 22

3. Research significance 24

CHAPTER II. Current robust design methods 25

A. Introduction 25

B. Robust design models 25

1. Two-step model 27

2. Dual response model 28

3. MSE Model 29

4. Weighted-sum model 29

5. QLF Model 30

6. Desirability function model 30

7. Prioritized model 31

8 Bias specified RD model 31

9. Variability specified RD model 32

10. Weighted-tchebycheff based model 32

11. Other robust design methods 33

CHAPTER III. Development of multidisciplinary pharmaceutical experimental design(PED) model 34

A. Development of multidisciplinary pharmaceutical experimental design(PED) model 34

1. Introduction 34

2. Development of the proposed PED model 36

a. Development of a response surface simplex design (RSSD) model 36

b. Development of a priority-based multidisciplinary RD model 37

3. Numerical example 39

4. Conclusions 42

CHAPTER IV. Development of robust data mining procedure (RDMP) 43

A. Development of robust data mining procedure (RDMP) 43

1. Introduction 43

2. Development of the proposed RDMP 45

a. Expectation-maximization (EM) algorithm 45

b. Correlation-based feature selection (CBFS) method 47

i. Best first search (BFS) algorithm 49

c. Connection to multidisciplinary RD 50

i. Response surface methodology (RSM) 50

ii. Development of a robust desirability function (RDF) model 51

iii. Development of a priority-based multidisciplinary RD model 53

3. Numerical example 54

a. Stage I : Em algorithm for data pre-treatment 56

b. Stage II: Data mining for dimensionality reduction 56

c. Stage III: The results of multidisciplinary RD using RSM 57

4. Conclusions 59

CHAPTER V. Conclusions and further research 61

A. Conclusions 61

B. Further research 62

References 64

Published or submitted papers 68

Published or submitted conference(comference) proceedings 69

Appendixes 70

Appendix III. Matlab codes 70

Appendix IV. Matlab codes 72

LIST OF TABLES

Table 1. Experimental format 17

Table 2. Summary of robust design (RD) models 26

Table 3. Experimental format of Taguchi's approach to RD 28

Table 4. The quality characteristic of friability (F), hardness (H), and disintegration (D) 40

Table 5. Design matrix 40

Table 6. The optimal solution of the tablet manufacturing process 42

Table 7. The quality characteristic of friability (F), hardness (H), and disintegration (D) 55

Table 8. The data set for the tablet manufacturing process 56

Table 9. Estimated mean, standard deviation, and value through 25 times iteration of the EM algorithm 56

Table 10. DM results for response y₁, y₂, and y₃ 57

Table 11. The optimal solution of the tablet manufacturing process using the RDF model 59

Table 12. The optimal solution of the tablet manufacturing process using the priority-based multidisciplinary RD model 59

LIST OF FIGURES

Figure 1. Taguchi's quality design 16

Figure 2. Comparison of simplex-lattice design for (a)k=3 m=2, (b)k=3 m=3, (c)and simplex-centroid design for k=3 18

Figure 3. Process distribution functions 20

Figure 4. The tablet manufacturing process 22

Figure 5. An overview of the research framework 23

Figure 6. An overview of the proposed multidisciplinary PED model 36

Figure 7. Ternary plot 40

Figure 8. An overview of the proposed RDMP 45

Figure 9. An overview of EM algorithm for data pre-treatment 47

Figure 10. An overview of CBFS method for dimensionality reduction 50

Figure 11. Hierarchical multivariate RD optimization 63

Figure 12. Integrated experimental design 63