Entry requirements

120 credits including 90 credits in statistics and 7.5 credits programming in R, Python or Julia.

Learning outcomes

After completing the course, the student is expected to:

• have a good knowledge of a large number of machine learning models
• be able to use methods for evaluating and improving predictive models
• be able to describe and discuss ethical aspects of big data and black box-models
• be able to handle big data
• be able to train and use machine learning models in R
• be able to train and use neural networks using Keras/Tensorflow

Content

Regularised regression, nearest neighbour methods, decision trees, ensemble models, bagging, out-of-sample evaluations, handling of big data, ethical questions regarding big data and predictive models, methods for explainable machine learning, and neural networks: architectures, gradient descent, generative models, regularisation and adversarial examples.

Instruction

Instruction is given in the form of lectures, labs and/or as seminars.

Assessment

The examination takes place through written and/or oral presentation of compulsory assignments.

Other directives

This course is part of the master degree program in statistics.