Foundations of machine learning

Aim

The aim of the course is to explain what machine learning is, what types of machine learning that exist, their possibilities and limitations, and to get an overview of the different methods and techniques used for machine learning. After the course, the student will be able to:

• Describe what machine learning can be used for, as well as its possibilities and limitations.
• Describe how data is used to train a machine learning model to solve a given problem.
• Describe the conditions under which machine learning methods can be expected to perform well.
• Formulate machine learning as optimization problems.
• Distinguish between supervised and unsupervised machine learning.
• Evaluate the performance of a machine learning model.
• Account for common methods and models used in machine learning.

Content

The course consists of seven parts:

• What is machine learning?
• Supervised learning: a first approach
• Parametric models and a statistical perspective on learning
• Understanding, evaluating, and improving performance
• Neural networks and deep learning
• Generative models and learning from unlabeled data
• User aspects of machine learning