Muutke küpsiste eelistusi

Optimal Design of Experiments: A Case Study Approach [Kõva köide]

4.58/5 (33 hinnangut Goodreads-ist)
(University of Antwerp and Erasmus), (JMP Division of SAS, USA)
  • Formaat: Hardback, 304 pages, kõrgus x laius x paksus: 234x158x22 mm, kaal: 553 g
  • Ilmumisaeg: 01-Jul-2011
  • Kirjastus: John Wiley & Sons Inc
  • ISBN-10: 0470744618
  • ISBN-13: 9780470744611
Teised raamatud teemal:
Optimal Design of Experiments: A Case Study ApproachSuurem pilt
  • Formaat: Hardback, 304 pages, kõrgus x laius x paksus: 234x158x22 mm, kaal: 553 g
  • Ilmumisaeg: 01-Jul-2011
  • Kirjastus: John Wiley & Sons Inc
  • ISBN-10: 0470744618
  • ISBN-13: 9780470744611
Teised raamatud teemal:
Püsilink: https://www.kriso.ee/db/9780470744611.html
Märksõnad:
This book demonstrates the utility of the computer-aided optimal design approach using real industrial examples-- This book demonstrates the utility of the computer-aided optimal design approach using real industrial examples. These examples address questions such as the following: How can I do screening inexpensively if I have dozens of factors to investigate? What can I do if I have day-to-day variability and I can only perform 3 runs a day? How can I do RSM cost effectively if I have categorical factors? How can I design and analyze experiments when there is a factor that can only be changed a few times over thestudy? How can I include both ingredients in a mixture and processing factors in the same study? How can I design an experiment if there are many factor combinations that are impossible to run? How can I make sure that a time trend due to warming up of equipment does not affect the conclusions from a study? How can I take into account batch information in when designing experiments involving multiple batches? How can I add runs to a botched experiment to resolve ambiguities While answering these questions the book also shows how to evaluate and compare designs. This allows researchers to make sensible trade-offs between the cost of experimentation and the amount of information they obtain. The structure of the book is organized around the following chapters: 1) Introduction explaining the concept of tailored DOE. 2) Basics of optimal design. 3) Nine case studies dealing with the above questions using the flow: description → design → analysis → optimization or engineering interpretation. 4)Summary. 5) Technical appendices for the mathematically curious-- Provided by publisher. This book demonstrates the utility of the computer-aided optimal design approach using real industrial examples. These examples address questions such as the following:• How can I do screening inexpensively if I have dozens of factors to investigate?• What can I do if I have day-to-day variability and I can only perform 3 runs a day?• How can I do RSM cost effectively if I have categorical factors?• How can I design and analyze experiments when there is a factor that can only be changed a few times over the study?• How can I include both ingredients in a mixture and processing factors in the same study?• How can I design an experiment if there are many factor combinations that are impossible to run?• How can I make sure that a time trend due to warming up of equipment does not affect the conclusions from a study?• How can I take into account batch information in when designing experiments involving multiple batches?• How can I add runs to a botched experiment to resolve ambiguities?While answering these questions the book also shows how to evaluate and compare designs. This allows researchers to make sensible trade-offs between the cost of experimentation and the amount of information they obtain. The structure of the book is organized around the following chapters: 1) Introduction explaining the concept of tailored DOE. 2) Basics of optimal design. 3) Nine case studies dealing with the above questions using the flow:• description → design → analysis → optimization or engineering interpretation. 4) Summary.5) Technical appendices for the mathematically curious.
Preface xiii
Acknowledgments xv
1 A simple comparative experiment
1(8)
1.1 Key concepts
1(1)
1.2 The setup of a comparative experiment
2(6)
1.3 Summary
8(1)
2 An optimal screening experiment
9(38)
2.1 Key concepts
9(1)
2.2 Case: an extraction experiment
10(11)
2.2.1 Problem and design
10(4)
2.2.2 Data analysis
14(7)
2.3 Peek into the black box
21(23)
2.3.1 Main-effects models
21(1)
2.3.2 Models with two-factor interaction effects
22(2)
2.3.3 Factor scaling
24(1)
2.3.4 Ordinary least squares estimation
24(3)
2.3.5 Significance tests and statistical power calculations
27(1)
2.3.6 Variance inflation
28(1)
2.3.7 Aliasing
29(4)
2.3.8 Optimal design
33(2)
2.3.9 Generating optimal experimental designs
35(5)
2.3.10 The extraction experiment revisited
40(1)
2.3.11 Principles of successful screening: sparsity, hierarchy, and heredity
41(3)
2.4 Background reading
44(1)
2.4.1 Screening
44(1)
2.4.2 Algorithms for finding optimal designs
44(1)
2.5 Summary
45(2)
3 Adding runs to a screening experiment
47(22)
3.1 Key concepts
47(1)
3.2 Case: an augmented extraction experiment
48(11)
3.2.1 Problem and design
48(7)
3.2.2 Data analysis
55(4)
3.3 Peek into the black box
59(8)
3.3.1 Optimal selection of a follow-up design
60(5)
3.3.2 Design construction algorithm
65(1)
3.3.3 Foldover designs
66(1)
3.4 Background reading
67(1)
3.5 Summary
67(2)
4 A response surface design with a categorical factor
69(26)
4.1 Key concepts
69(1)
4.2 Case: a robust and optimal process experiment
70(12)
4.2.1 Problem and design
70(9)
4.2.2 Data analysis
79(3)
4.3 Peek into the black box
82(10)
4.3.1 Quadratic effects
82(1)
4.3.2 Dummy variables for multilevel categorical factors
83(3)
4.3.3 Computing D-efficiencies
86(1)
4.3.4 Constructing Fraction of Design Space plots
87(1)
4.3.5 Calculating the average relative variance of prediction
88(2)
4.3.6 Computing I-efficiencies
90(1)
4.3.7 Ensuring the validity of inference based on ordinary least squares
90(1)
4.3.8 Design regions
91(1)
4.4 Background reading
92(1)
4.5 Summary
93(2)
5 A response surface design in an irregularly shaped design region
95(18)
5.1 Key concepts
95(1)
5.2 Case: the yield maximization experiment
95(13)
5.2.1 Problem and design
95(8)
5.2.2 Data analysis
103(5)
5.3 Peek into the black box
108(4)
5.3.1 Cubic factor effects
108(1)
5.3.2 Lack-of-fit test
109(2)
5.3.3 Incorporating factor constraints in the design construction algorithm
111(1)
5.4 Background reading
112(1)
5.5 Summary
112(1)
6 A "mixture" experiment with process variables
113(22)
6.1 Key concepts
113(1)
6.2 Case: the rolling mill experiment
114(9)
6.2.1 Problem and design
114(7)
6.2.2 Data analysis
121(2)
6.3 Peek into the black box
123(9)
6.3.1 The mixture constraint
123(1)
6.3.2 The effect of the mixture constraint on the model
123(2)
6.3.3 Commonly used models for data from mixture experiments
125(2)
6.3.4 Optimal designs for mixture experiments
127(3)
6.3.5 Design construction algorithms for mixture experiments
130(2)
6.4 Background reading
132(1)
6.5 Summary
133(2)
7 A response surface design in blocks
135(28)
7.1 Key concepts
135(1)
7.2 Case: the pastry dough experiment
136(15)
7.2.1 Problem and design
136(8)
7.2.2 Data analysis
144(7)
7.3 Peek into the black box
151(9)
7.3.1 Model
151(2)
7.3.2 Generalized least squares estimation
153(3)
7.3.3 Estimation of variance components
156(1)
7.3.4 Significance tests
157(1)
7.3.5 Optimal design of blocked experiments
157(1)
7.3.6 Orthogonal blocking
158(2)
7.3.7 Optimal versus orthogonal blocking
160(1)
7.4 Background reading
160(1)
7.5 Summary
161(2)
8 A screening experiment in blocks
163(24)
8.1 Key concepts
163(1)
8.2 Case: the stability improvement experiment
164(15)
8.2.1 Problem and design
164(5)
8.2.2 Afterthoughts about the design problem
169(6)
8.2.3 Data analysis
175(4)
8.3 Peek into the black box
179(5)
8.3.1 Models involving block effects
179(3)
8.3.2 Fixed block effects
182(2)
8.4 Background reading
184(1)
8.5 Summary
185(2)
9 Experimental design in the presence of covariates
187(32)
9.1 Key concepts
187(1)
9.2 Case: the polypropylene experiment
188(18)
9.2.1 Problem and design
188(9)
9.2.2 Data analysis
197(9)
9.3 Peek into the black box
206(10)
9.3.1 Covariates or concomitant variables
206(1)
9.3.2 Models and design criteria in the presence of covariates
206(5)
9.3.3 Designs robust to time trends
211(4)
9.3.4 Design construction algorithms
215(1)
9.3.5 To randomize or not to randomize
215(1)
9.3.6 Final thoughts
216(1)
9.4 Background reading
216(1)
9.5 Summary
217(2)
10 A split-plot design
219(36)
10.1 Key concepts
219(1)
10.2 Case: the wind tunnel experiment
220(20)
10.2.1 Problem and design
220(12)
10.2.2 Data analysis
232(8)
10.3 Peek into the black box
240(13)
10.3.1 Split-plot terminology
240(2)
10.3.2 Model
242(2)
10.3.3 Inference from a split-plot design
244(3)
10.3.4 Disguises of a split-plot design
247(2)
10.3.5 Required number of whole plots and runs
249(1)
10.3.6 Optimal design of split-plot experiments
250(1)
10.3.7 A design construction algorithm for optimal split-plot designs
251(2)
10.3.8 Difficulties when analyzing data from split-plot experiments
253(1)
10.4 Background reading
253(1)
10.5 Summary
254(1)
11 A two-way split-plot design
255(22)
11.1 Key concepts
255(1)
11.2 Case: the battery cell experiment
255(12)
11.2.1 Problem and design
255(8)
11.2.2 Data analysis
263(4)
11.3 Peek into the black box
267(8)
11.3.1 The two-way split-plot model
269(1)
11.3.2 Generalized least squares estimation
270(3)
11.3.3 Optimal design of two-way split-plot experiments
273(1)
11.3.4 A design construction algorithm for D-optimal two-way split-plot designs
273(1)
11.3.5 Extensions and related designs
274(1)
11.4 Background reading
275(1)
11.5 Summary
276(1)
Bibliography 277(6)
Index 283
Peter Goos, Department of Mathematics, Statistics and Actuarial Sciences of the Faculty of Applied Economics of the University of Antwerp. His main research topic is the optimal design of experiments. He has published a book as well as several methodological articles on the design and analysis of blocked and split-plot experiments. Other interests of his in this area include discrete choice experiments, model-robust designs, experimental design for non-linear models and for multiresponse data, and Taguchi experiments. He is also a member of the editorial review board of the Journal of Quality Technology. Bradley Jones, Senior Manager, Statistical Research and Development in the JMP division of SAS, where he leads the development of design of experiments (DOE) capabilities in JMP software. Dr. Jones is widely published on DOE in research journals and the trade press. His current interest areas are design of experiments, PLS, computer aided statistical pedagogy, and graphical user interface design.