Contact
Ludwig-Maximilians-Universität München
Lehrstuhl für Datenbanksysteme und Data Mining
Oettingenstraße 67
80538 München
Germany
Lehrstuhl für Datenbanksysteme und Data Mining
Oettingenstraße 67
80538 München
Germany
Room:
F 110
Phone:
+49-89-2180-9304
Email:
leiber@dbs.ifi.lmu.de
Teaching
- Datenbanksysteme [ WiSe19/20, WiSe20/21, WiSe21/22, WiSe22/23, WiSe23/24 ]
- Einführung in die Informatik: Systeme und Anwendungen [ SoSe20, SoSe21, SoSe22, SoSe23 ]
- Bachelorseminar: Information-theoretic Data Mining [ WiSe19/20, SoSe20, SoSe21 ]
- Bachelorseminar: Data Mining [ WiSe21/22, SoSe22, WiSe22/23, SoSe23, WiSe23/24 ]
- Oberseminar (Prof. Böhm) [ WiSe19/20, SoSe20, WiSe20/21, SoSe21, WiSe21/22, SoSe22, WiSe22/23, SoSe23, WiSe23/24 ]
Research Interests
- Estimating the Number of Clusters
- Hartigan's Dip-Test of Unimodality
- Deep Clustering
- (Common) Subspace Clustering
- Non-Redundant/Alternative Clustering
- Information Theoretic Clustering
Supervised Theses
Bachelor
- Numerical Algorithms for Symmetric Eigenvalue Decomposition in Subspace k-means Clustering
- Density-Based Clustering: A Robust Multi-Density Alternative to DBSCAN
- Improving the FOSSCLU Algorithm:Optimized Parameter-search and Clustering in Non-redundant Subspaces
- Ein nicht-parametrischer Ansatz zur Erkennung beliebiger Clusterstrukturen
- Automated tumor detection based on routine blood tests using Artificial Neural Networks
- Improving Non-Redundant Clustering Approaches by K-means Extensions
- Acquiring Meaningful Data Representations by Combining Autoencoders and Hartigan's Dip Test
- Parameterfreies nicht-redundantes Clustering mithilfe des Dip-Tests
- Self-Supervised Deep Clustering for Tabular Data
- Design and Implementation of a Database with Web Interface for Side Effects of Immunotherapies
- Evaluation of Density Based Mode Seeking
- MDL-based Parameter Selection for Subspace and Non-Redundant Clustering
Master
- Informationsextraktion aus Einverständniserklärungen von klinischen Studien mittels Natural Language Processing und Maschinellem Lernen
- Deep Learning-based Forensic Age Estimation from CT Images
- Eine Prognose der COVID-19-Fallzahlen in Deutschland durch Neuronale Netze basierend auf Veränderungen der Mobilität
- Minimal Description Length Principle for Parameter-free Motif Detection in Time Series Data
- k-SubMix: Clustering on mixed-type data in optimal Subspaces
- Subspace Clustering nicht-konvexer Strukturen durch Micro-Cluster Graphoptimierung
- A Survey of Unimodality Tests and their Potential in Clustering Environments
- Uncertainty Quantification in Deep Learning-Based Forensic Age Estimation
- Deep Density-Based Clustering using Contrastive Learning
- Building Decision Trees using Hartigans Dip-Test of Unimodality
Publications
- Dip-based Deep Embedded Clustering with k-Estimation
- Automatic Parameter Selection for Non-Redundant ClusteringProceedings of the 2022 SIAM International Conference on Data Mining (SDM). Society for Industrial and Applied Mathematics, 2022. [DOI][Supplement][Code]
- The DipEncoder: Enforcing Multimodality in Autoencoders
- Extension of the Dip-test Repertoire - Efficient and Differentiable p-value Calculation for ClusteringProceedings of the 2023 SIAM International Conference on Data Mining (SDM). Society for Industrial and Applied Mathematics, 2023. [DOI][Code + Supplement]
- k-SubMix: Common Subspace Clustering on Mixed-Type Data
- Non-Redundant Image Clustering of Early Medieval Glass Beads
- Application of Deep Clustering AlgorithmsProceedings of the 32nd ACM International Conference on Information and Knowledge Management, 2023. [DOI]