Multivariate Data Analysis and Experimental Design
5 credits
Syllabus, Bachelor's level, 1MB344
A revised version of the syllabus is available.
- Code
- 1MB344
- Education cycle
- First cycle
- Main field(s) of study and in-depth level
- Computer Science G2F, Technology G2F
- Grading system
- Fail (U), Pass (3), Pass with credit (4), Pass with distinction (5)
- Finalised by
- The Faculty Board of Science and Technology, 25 April 2017
- Responsible department
- Biology Education Centre
Entry requirements
60 credits within the programme Degree of Master of Science in Engineering including Scientific Computing I, Probability and Statistics, and Linear Algebra II. Completed Programming Techniques II
Learning outcomes
After successful completion of the course the student should be able to
- use and theoretically describe some of the basic methods for exploratory multivariate data analysis; data compression and visualisation using principal-component analysis and cluster analysis .
- use and theoretically describe some of the basic methods for predictive multivariate data analysis; conventional least square adjustments for linear models and design of linear discriminants for classification.
- use basic experimental design methods; factorial design and respons surface designe in two levels.
- carry out and implement the analyses and interpretations using at least one general software environment for the analyses, for example MATLAB or R.
Content
Exploratory multivariate data analysis: principal component analysis, clustering etc. Predictive multivariate data analyses: model based and model-free linear/nonlinear classification and regression, model selection and performance estimation. Applications.
Instruction
Lectures, seminars, exercises based on manual as well as computer calculations.
Assessment
Computer exercises, assignments and written exam.