The teaching is organized as follows:

Learning outcomes

The course is aimed at giving theoretical and applicative fundamentals of Pattern Recognition, a class of methodologies usable to recognize and retrieve information from biological data.
In particular all the fundamental aspects of PR will be presented: data representation, classification, clustering and validation. The main focus will be more on the methodologies rather than on the application programs (already seen in other courses)

The first part of the course will present, from a general perspective, methods, motivations and problems underlying the main techniques employed in Pattern Recognition. The second part, more devoted to applications, will describe practical bioinformatics scenarios where PR tools are applied (gene expression data analysis, protein remote homology detection, image segmentation and so on)
A final part will be devoted to practical implementation, via the MATLAB language, of some of the techniques studied in the previous parts.

Program

- General introduction to Pattern Recognition
- Data representation
- Bayes decision theory
- Generative and discriminative classifiers
- Validation of classification
- Neural Networks
- Hidden Markov Models
- Dissimilarities
- Clustering approaches
- Validation of clustering
- Applications

Examination Methods

Seminar on a topic related to Pattern Recognition and Bioinformatics
Written exam on topics treated during frontal lessons