Learning Outcomes:
On completion of this module, students should be able to:
1. Understand the fundamental concepts and principles of PLS, PCR, ILS and CLS regression.
2. Apply classification techniques to multivariate data using 2 class, 1 class or multiclass classifiers.
3. Analyse, optimise and validate regression and classification models
4. Apply variable selection to identify and interpret the most important variables in a regression or classification problem.
5. Evaluate a selection of chemometric techniques to achieve a pre-defined goal.
Indicative Module Content:
1. Mathematical basis, applications and limitations of univariate & multivariate calibration
2. Classical Least Squares, Inverse Lease Squares, Multivariate linear regression with interaction terms
3. Mathematical basis, applications and limitations of Ridge regression, principal components regression and partial least squares regression
4. Outlier evaluation, detection and removal
5. Cross validation for model building and optimisation
6. Mathematical basis & application of 1, 2 and multi-class classification techniques (e.g. LDA, QDA, PLS-DA, SIMCA)
7. Commonly used methods of variable selection
8. Application of variable selection to multivariate data and interpretation the results
9. Analysis of multivariate data from diverse sources
10. Development of MATLAB scripts for full dataflow