|
|
|||||||||||||||||
Course Catalog 2012-2013
SGN-2556 Pattern Recognition, 5 cr |
Additional information
Suitable for postgraduate studies
Person responsible
Juho Vihonen, Ari Visa, Joni Kämäräinen, Jussi Tohka, Ulla Ruotsalainen
Lessons
| Study type | P1 | P2 | P3 | P4 | Summer | Implementations | Lecture times and places |
|
|
|
|
|
|
|
|
|
Requirements
Exam and Matlab exercises. The exercises are mandatory.
Completion parts must belong to the same implementation
Principles and baselines related to teaching and learning
-
Learning outcomes
The aim is to deepen the understanding of pattern recognition principles and give students some ability to apply the methods on real problems. The aim is also to learn how to write in a scientific publication about the methods and the pattern classification results.
Content
| Content | Core content | Complementary knowledge | Specialist knowledge |
| 1. | Bayesian decision theory and Bayesian parameter estimation. Belief networks (graphical models). Linear and generalized linear discriminant functions | Hidden Markov models | |
| 2. | Stochastic pattern classification methods; Boltzman learning, Evolutionary methods in optimization | Genetic programming | |
| 3. | Nonmetric classification methods: CART, tree methods in principle. | Grammatical methods | |
| 4. | Algorithm-independent machine learning. No free lunch theorem, cross-validation, bootstrap, bagging and boosting | ||
| 5. | Unsupervised learning and clustering, fuzzy clustering methods, component analysis methods.Mixture densities, Hierarchical clustering, PCA and ICA | on-line clustering, graph theoretic methods, |
Evaluation criteria for the course
In order to pass the course the student has to complete all the exercises and get half of the maximum points from the exam. Grading is pass/fail.
Assessment scale:
Numerical evaluation scale (1-5) will be used on the course
Partial passing:
Study material
| Type | Name | Author | ISBN | URL | Edition, availability, ... | Examination material | Language |
| Book | "Pattern Classification" | Duda RO, Hart PE, Stork DG | 2nd edition, Wiley, 2001 | English |
Prerequisites
| Course | Mandatory/Advisable | Description |
| SGN-2500 Johdatus hahmontunnistukseen | Mandatory | 1 |
| SGN-2506 Introduction to Pattern Recognition | Mandatory | 1 |
1 . Either course SGN-2500 or SGN-2506 or corresponding knowledge is required
Additional information about prerequisites
Good Matlab programming skills are required;
Experience in academic/scientific writing in English is required (e.g. Bachelor or Master thesis)
Prerequisite relations (Requires logging in to POP)
Correspondence of content
| Course | Corresponds course | Description |
|
|
|
|
|
|
|
More precise information per implementation
| Implementation | Description | Methods of instruction | Implementation |
| Postgraduate course on pattern recognition to deepen the knowledge of pattern recognition methods. The aim of the course is to provide ability to apply the methods in student's own research work. The Matlab exercises are essential part of the course giving the possibility to utilize the methods in practical problems. |