Organization
- Course
- 3+2 hours weekly (equals 6 ECTS)
- Lecture:
- Prof. Dr. Thomas Seidl
- Assistant:
- Mamdouh Ajoud
- Audience:
- Master students in the programs of the Institute for Informatics
- Course Material:
- DeepClustering@Moodle
- Prior Knowledge:
- The course expects participants to have basic skills in machine learning and data mining.
- Course Language:
- English
Content
Clustering is a key unsupervised learning technique used to group data points based on similarity, revealing hidden patterns without prior labels. Traditional clustering methods often struggle in high-dimensional spaces, which led to the development of specialized approaches such as subspace clustering and deep clustering. Subspace clustering identifies clusters within relevant subsets of features, allowing for meaningful grouping even when noise or irrelevant dimensions exist. Deep clustering, on the other hand, integrates deep neural networks with clustering objectives to learn robust feature representations and complex, non-linear cluster structures. Together, these methods expand the applicability of clustering to modern, high-dimensional, and unstructured data domains.