This minicourse which is meant to introduce PhD students of the School of Agriculture, Forestry, Food and Environmental Sciences to multivariate statistical techniques. The course was held in the past at University College Cork (Cork, IRL) from May 2, 2012 to May 4, 2012 within the framework of a LLP-Erasmus agreement between UCC and Università degli Studi della Basilicata. This is a slightly larger version, approx 3 ECTS credits
Course objective: to provide an introduction to basic multivariate statistical analysis techniques in food science and nutrition
Prerequisites: basic knowledge of statistics (variables, hypothesis testing, univariate techniques). Ability to use a statistical analysis software.
Lectures (8×2 h). Introduction. Planning statistical analysis. An overview of descriptive and inferential multivariate statistical analysis techniques. Bad data and good data. Multivariate data. Preprocessing, transformation and standardization. Know thy batch: graphic exploration of multivariate data. Principal component analysis. Multidimentional scaling. Hierarchical and non-hierarchical cluster analysis. Neural networks. Principal Component Regression and PLS regression. Discriminant analysis.
Workshop (4×2 h). Individual and group work: statistical analysis of data in food science, technology and nutrition.
Please note: this course will be delivered in two modes
a. in a classroom (ASD, 5th floor building 3B) for Master and PhD students of
b. over the internet, using vyew (http://vyew.com). Addresses for the virtual rooms will be provided, but to interact with the lecturer an E-mail registration (firstname.lastname@example.org) is needed. Only 20 remote students can be accepted.
The first introductory lecture is scheduled for December 10th, 16.30 Rome time in room http://vyew.com/room#/663583/Multistatlecture1
- 10/12/2012 16:30-18:30 Introduction to multivariate statistics
- 13/12/2012 16:30-18:30 Graphic exploration of multivariate data. Statistical software
- 17/12/2012 16:30-18:30 Principal Component Analysis
- 10/01/2013 16:30-18:30 Multidimensional scaling. Multivariate plots
- 14/01/2013 16:30-18:30 Hierarchical and non-hierarchical cluster analysis
- 17/01/2013 16:30-18:30 Workshop on PCA, MDS, CA
- 21/01/2013 16:30-18:30 Artificial neural networks. Discriminant analysis. Classification and regression trees
- 24/01/2013 16:30-18:30 Principal Component Regression. Partial Least Square Regression
- 28/01/2013 16:30-18:30 Workshop on inferential methods
- 31/01/2013 16:30-18:30 Workshop
- Introductory textbooks:
- Gacula, M., Singh, J., Bi, J., Altan, S. 2008. Statistical methods in food and consumer research. Academic Press.
- Multivariate techniques:
- Arvanitoyannis, I.S. and Tzouros N.E. 2005. Implementation of quality control methods in conjunction with chemometrics toward authentication of dairy products. Critical Reviews in Food Science and Nutrition 45: 231–249.
- Everitt B.S., Dunn, D. 2001. Applied multivariate data analysis. Arnold
- Everitt B.S., Landau S., Leese M. 2001. Cluster analysis. Arnold
- Parente E. 2011. Analytical methods: Multivariate statistical tools for analytical data. In: Encyclopedia of Dairy Science, 2nd edition, John W. Fuquay, editor in chief. Elsevier. ISBN 978-0-12-374402-9, pp 93-102
- Wold S., Sjöström M., Eriksson L. 2001. PLS-regression: a basic tool of chemometrics. Chemometrics and Intelligent Laboratory Systems 58: 109–130
- Lecture 1: Introduction
- Lecture 2: Exploring multivariate data
- Lecture 3: Principal Component Analysis
- Lecture 4: Multidimensional scaling
- Lecture 5.1: Introduction to cluster analysis
- Lecture 5.2: Hierarchical and non-hierarchical cluser analysis
- Lecture 6: Artificial Neural Networks
- Lecture 7: Multivariate regression problems
- Lecture 8: Supervised pattern recognition