Applied Multivariate Statistical Analysis (B-KUL-I0P16B)
Aims
Present the concepts and methods of multivariate analysis, emphasizing the applications and attempting to make the mathematics as palatable as possible.
The student is expected to:
- Apply linear algebra in variance, covariance and correlation structures and understand geometrical equivalents of basic multivariate reasoning
- Understand properties and applications of the Multivariate Normal distribution
- Carry out inference about multivariate means
- Understand and apply basic ordination, discrimination and classification methodologies: Principal Components Analysis, Factor Analysis, Discriminant Analysis and Cluster Analysis
- Be able to apply these methods on real datasets
- Make use of existing software packages to solve problems in Multivariate Analysis
Previous knowledge
Calculus, Linear algebra, introductory statistics, linear models.
Order of Enrolment
This course unit is a prerequisite for taking the following course units:
I0U20A : Integrated Bioinformatics Project
Is included in these courses of study
- Master in de bio-ingenieurswetenschappen: biosysteemtechniek (Leuven) (Major technologie voor de agrovoedingssector) 120 ects.
- Master of Bioinformatics (Leuven) 120 ects.
- Master in de bio-ingenieurswetenschappen: landbouwkunde (Leuven) 120 ects.
- Master in de bio-informatica (Leuven) 120 ects.
- Master of Statistics and Data Science (on campus) (Leuven) (Statistics and Data Science for Biometrics) 120 ects.
- Courses for Exchange Students Faculty of Bioscience Engineering (Leuven)
- Master of Bioscience Engineering: Agro- and Ecosystems Engineering (Leuven) (Major Subject: Agricultural and Resource Economics) 120 ects.
- Master of Bioscience Engineering: Agro- and Ecosystems Engineering (Leuven) (Major Subject: Plant Production Systems) 120 ects.
Activities
4 ects. Applied Multivariate Statistical Analysis (B-KUL-I0P16a)
Content
Introduction: overview of different Multivariate Analysis concepts and methodologies, review of linear algebra, the Multivariate Normal distribution, sample geometry, centering and scaling, exploratory versus confirmatory analysis, geometric equivalences in multivariate analysis, multivariate data visualisation.
The core of the course consists of two major parts:
1. Ordination methods or the analysis of covariance structures
- Principal Components Analysis, simple Correspondence Analysis
- Factor Analysis
- Biplotting
- Partial Least Squares
- Multidimensional Scaling
2. Parametric and non-parametric classification methods
- Hierarchical and non-hierarchical Cluster Analysis, one-dimensional projection pursuit fo clustering data
- Discriminant analysis, Tree Based Models, logistic regression
Course material
Slides, knowledge clips
Format: more information
Live or online blended, with knowledge clips
1 ects. Applied Multivariate Statistical Analysis: Practical Exercises (B-KUL-I0P17a)
Content
R demo-programs of all multivariate methodologies of the lectures. Discussion and R-analysis of real-world data science problems. Take-home problems. Paper development, including problem and data description, data management, R-programming and analysis, interpretation, conclusion.
Course material
R-programs, take-home problems, datasets, knowledge clips.
Format: more information
Live or online blended with knowledge clips
Evaluation
Evaluation: Applied Multivariate Statistical Analysis (B-KUL-I2P16b)
Explanation
Examination (open book):
- 4 questions
- 1 question about the paper
The paper counts for 1/4 of the final score for the course.
Information about retaking exams
Students who did not obtain a sufficient score for the paper, need to retake the paper during the third examination period.
In case students obtained a sufficient score for the paper in the first examination period, this score will be maintained in the third examination period.