Pre requisite: Basic Probability and Statistics
Brief syllabus: Graphical data analysis tools, testing a new method and comparing two or more methods. Deficiencies of one factor at a time experiments, necessity of randomization. Two level factorials, judging significance in unreplicated and replicated experiments, developing mathematical models, whether experiment is satisfactory. Checking if linear model sufficient, dealing with large number of factors, Determining optimum conditions through experimentation, Usage of software, project.
Detailed Syllabus:
Module I:
Graphical Data analysis tools - Stem and leaf plot, Dot plot, Box plot
Distribution of sample data - Normal distribution, t distribution
Normal Probability Plotting on ordinary graph paper, Interpretation
Testing a new method for improvement - Variability known from past, Variability estimated from the experiment
Comparing two methods - Randomised samples, Paired samples
Comparing more than two methods simultaneously- ANOVA
Module II:
Experimental strategies - Deficiencies of one factor at a time experiments, Problems in analysis of past data, Necessity for randomization
Basics of Experimental Design - Terminology, Two level factorials, Estimation of effects and interactions, Yates algorithm, Unreplicated experiments - judging significance, Testing for significance in replicated experiments.
Developing mathematical model equations, calculating residuals, checking whether experiment has been conducted satisfactorily.
Module III:
Handling non-normal response - Transformations.
Choosing the number of experiments, Testing whether linear model is satisfactory, How to handle uncontrollable factors, How to deal with difficult to randomise factors.
Dealing with large number of factors, Fractional Factorial experiments and Plackett Burman Designs, How to minimise possible confusion, Design Resolution, Sequential experimentation strategies, Folding over.
Module IV:
Determining optimum conditions experimentally - Central Composite Designs, D-Optimal Designs, Response Surface methods, Mixture experiments.
Experiments to determine variability and minimise it.
Training in Design Expert, software for DoE. Individual Design Project, presentation and discussion. Applications / Case Studies in Research, Quality Improvement, Product Development.
Textbook: Anderson, Mark J., & Whitcomb, Patrick J., "Doe Simplified: Practical Tools for Effective Experimentation", Productivity Press, USA.
References:
HOME | NOTES MAIN | LEVEL ABOVE
© www.dhanish.150m.com