WIAS/PE&RC course Statistics for Data Science

Background

In many areas of science new tools and strategies are developed to measure multiple features at subjects and objects of interest. Typical for these new types of data is that they occur in large volumes, are high-dimensional, or are hierarchically structured. For example, in genetics (humans, animals, plants) data can be available at the levels of DNA, RNA, proteins, metabolites and all kinds of phenotypes. These new data require new techniques for analysis and visualization which are provided by a new science that combines elements of statistics and machine learning: Data Science.

Objectives

This course will make you familiar with a modern toolbox of analysis techniques at the interface of statistics and machine learning. You will develop the skills to build and evaluate modeling strategies for complex (big, high-dimensional, hierarchically-structured) data as occurring in the areas relevant to WIAS and PE&RC. Moreover, the course will give you insights in the connections between modern modeling strategies and will teach you to ask the right questions in order to choose the best method for the data at hand. Illustrations will come from relevant case studies.

General Information

Target Group	PhD candidates
Course level	In-depth, post graduate
Group size	A minimum of 10 and a maximum of 24 participants
Course duration	4 days and given yearly
Style of teaching	This will be a hands-on course in which each (interactive) lecture is followed by an illustrative practical.
Materials	- James, Witten, Hastie & Tibshirani (2021). An Introduction to Statistical Learning. Freely available here: https://hastie.su.domains/ISLR2/ISLRv2_website.pdf - R and RStudio.
Language	English
Credit points	1,2 ECTS
Name lecturers	Gerrit Gort, Carel Peeters and Jasper Engel (Biometrics, WUR)
Teaching methods	Lectures, computer practicals, discussions
In-person	Wageningen University campus, Wageningen, the Netherlands. Note: Room and building will be communicated to registered and confirmed participants about one week before the start of the course.

 

Programme

The course spans four days. The following topics will be discussed:

Day 1: Recap linear regression, resampling (bootstrap, jackknife, cross validation), regularized regression: ridge and lasso

Day 2: Extend linear model with splines, Generalized Additive Models (GAM)

Day 3: Classification: logistic regression and linear discriminant analysis; regression trees, random forests, boosting and bagging

Day 4: Clustering (hierarchical clustering, k-means), graphical modeling

Fee (includes coffee/tea and lunches)

Reduced fee: WUR PhD candidates with a TSP	€ 375
University fee: All other PhD candidates / post-docs and staff of Wageningen University	€ 750

Cancellation condition

You may cancel free of charge up to 1 month before the start of the course. After this date you will be charged with the University fee. Unless: 

• You can find someone to replace you in the course and supply the course coordinator with the name and contact information of your replacement. In this case you will be charged a €50,- cancellation fee.
• PhDs of Wageningen University who have a valid reason to cancel (illness or death in the family 1st or 2nd degree). In this case you will be charged the reduced fee and your supervisor/PI must send a mail indication the reason for cancellation.

More information: Please contact: WIAS Office wias@wur.nl