Pattern Recognition techniques help make sense of data. Unsupervised methods can help explore vast data sets. Supervised methods use training data to infer rules that allow making valid predictions. Applications abound in fields ranging from astronomy to finance.
This compact course aims to let you gain familiarity with some essential pattern recognition techniques. These are essential tools to address problems in prediction, tracking and exploratory data analysis. We will concentrate mostly on techniques that can be formulated as directed probabilistic graphical models, viz.:
- Logistic Regression
- Gaussian Mixture Models
- Hidden Markov Models
- Kalman Filter
- Topic Models
The course relies on basic notions of probability theory (joint distribution, marginal distribution). There will be a brief reminder, but it helps to have studied these beforehand.
When and Where
Monday Oct. 7th -- Friday Oct. 11th 2013, 09:30 -- 12:30HCI, Speyerer Strasse 6, Second Floor.
For whom
Interested undergraduate and graduate students. The course amounts to 2 CP.If you plan to attend, please register by sending an email to fred.hamprecht@iwr.uni-heidelberg.de