Program

Machine Learning in Social Science Research

Methods and Applications

The course is organized into four thematic modules, one per day. Each day combines a lecture introducing core methods, an applied case study connecting the technique to a real social science problem, and from day two onwards a hands-on coding lab in R.

Time

Session

Instructor(s)

Day 1  Foundations: machine learning in social science    Tue 5 May

9:45 –10:15

Welcome and course overview

Tommaso Agasisti

10:15–11:00

Lecture: machine learning in social science: an overview

Ilja Cornelisz

11:00 –11:30

Coffee break

11:30–12:30

Lecture: regularized regression: LASSO, Ridge, and Elastic Net

Chris Van Klaveren

12:30–13:30

Lunch

13:30–14.30

Lecture: regularized regression: LASSO, Ridge, and Elastic Net

Chris Van Klaveren

14:30–15:15

Paper discussion: Mortality prediction; classification metrics and model validation

Ilja Cornelisz

15:15–15:30

Coffee break

 

15:30–16:15

Paper discussion: mortality prediction; classification metrics and model validation

Ilja Cornelisz

 

 

 

Day 2  Education: flexible models and support vector machines   Wed 6 May

09:00–11:00

Lecture: Flexible models: support vector machines and ensemble learning

Chris Van Klaveren

11:00–11.30

Coffee break

11:30–13:00

Paper discussion: Predicting student graduation; comparing SVM and regularized regression

Ilja Cornelisz

13:00–14:00

Lunch

14:00–16:00

R lab: Regularized regression and SVM for predicting students’ performance

Melisa Diaz Lema

 

 

 

Day 3  Criminal justice: tree-based models and random forests   Thu 7 May

09:00–11:00

Lecture: Tree-based models and random forests

Chris Van Klaveren

11:00–11:30

Coffee break

11:30–13:00

Paper discussion: Predicting recidivism using classification trees; ethical and social implications

Ilja Cornelisz

13:00–14:00

Lunch

14:00–14:30

Paper discussion: Predicting recidivism using classification trees; ethical and social implications

Ilja Cornelisz

14:30–16:00

R lab: Random forests for predicting students’ experiences of bullying

Melisa Diaz Lema

 

Day 4  Synthesis: neural networks and your own research   Fri 8 May

09:00–11:00

Lecture: Artificial neural networks: architecture, backpropagation, gradient descent

Chris Van Klaveren

11:00–11:30

Coffee break

11:30–13:00

Paper discussion: Medical outcome classification using neural networks

Ilja Cornelisz

13:00–14:00

Lunch

14:00–16:00

Research workshop and wrap-up: Applying ML methods to your own research project

All instructors

Times are indicative and may be adjusted on-site. All R labs assume basic familiarity with R; datasets and scripts will be provided in advance.