X. Model-based clustering techniques
Given by Sven Laur
Brief summary: Phylogenetic trees and hierarchical clustering. Independent mutation model. Hard clustering techniques. K-means clustering as a maximum likelihood estimate. Gaussian mixture model as a generalisation of k-means algorithm. Corresponding hard clustering algorithms.
Slides: PDF
Video: UTTV
Literature:
- Bishop: Pattern Recognition and Machine Learning pages 423 - 448
Complementary exercises:
- Bishop: Pattern Recognition and Machine Learning pages 455 - 459
- Practical clustering tasks:
- Discretisation of continuous variables, such as height and weight or gene expression
- Finding relations between such discretised signals
Free implementations: