Integrated Lecture "Cognitive Algorithms"

Computer programs can learn useful cognitive skills. This integrated lecture tries to communicate an intuitive understanding of elementary concepts in machine learning and their application on real data with a special focus on methods that are simple to implement. We recommend the "Machine Learning 1" lecture or the "Machine learning lab course" for a more advanced treatment (this course is not a prerequisite).

Dates

Lectures/Tutorials

Lectures and tutorials take place every other week respectively. There will be 3 groups for the tutorials, registration will happen via Moses. In the tutorials we will briefly recap the lecture and discuss more theoretical exercises.

Assignments

After each lecture there will be assignments where you mainly implement the algorithms discussed in the lecture. Those assignments can be handed in in groups which will be organised via ISIS.

Topics

We will cover (among other things)

• Supervised learning (linear regression techniques, linear classification, kernel based regression, neural networks)
• Unsupervised Learning (Principal Component Analysis, Clustering)
• Model Selection

More information can be found on the ISIS Website.

Language

English

Prerequisites

The following prerequisites are helpful for taking the course:

• Basic knowledge in linear algebra and calculus
• Basic programming knowledge, programming in Python

• To participate in the exam you first need to pass the elective course

Credits

The integrated lecture is the compulsory part of the B.Sc. module "Kognitive Algorithmen" in Computer Science. This "Kognitive Algorithmen" module is a 6 ECTS/SP module, and consists of

• The compulsory integrated lecture (2 SWS / 3 SP / 3 ECTS), and

• An elective (2 SWS / 3 SP / 3 ECTS): an introduction to python programming or a seminar ("Hot Topics in ML") for a more in-depth treatment of selected applications.

The grade will be determined in a written exam at the end of the semester. The grades of the elective will not count towards the grade of the entire module.