Uusi opinto-opas (sisältäen myös opetusohjelmat) lukuvuodelle 2018-2019 sijaitsee osoitteessa https://opas.peppi.utu.fi . Tältä sivustolta löytyvät enää vanhat opinto-oppaat ja opetusohjelmat.

The new study guide (incl. teaching schedules) for academic year 2018-2019 can be found at https://studyguide.utu.fi. This site contains only previous years' guides.

x !
Archived Curricula Guide 2013–2014
Curricula Guide is archieved. Please refer to current Curricula Guides
TKO_5519 Pattern Recognition 5 ECTS
Organised by
Computer Science
Person in charge
Tapio Pahikkala

Learning outcomes

Pattern recognition addresses the problems of identifying patterns from data and making decisions based on these patterns. This course aims at delivering an understanding of this discipline. After finishing the course, the student understands the core concepts of underfitting and overfitting, which are some of the central problems when applying pattern recognition algorithms for high dimensional data. The student also knows how to find a suitable balance between these extremes using
various different forms of regularization techniques. In addition, student is able to reliably estimate the prediction performance of pattern recognition methods and to recognize and avoid pitfalls causing biased estimates. Finally, the student can program a computationally
efficient method for a practical pattern recognition task.

Contents

This course includes introduction to the most important principles, algorithms, and practical solutions of pattern recognition. A special emphasis is also put on the problems of performance evaluation of pattern recognition methods with cross-validation, a concept of great importance in biomedical sciences, where amount of data available for
testing is often very low. As example pattern recognition methods, the course covers several variations of regularized least-squares, a reasonably simple but surprisingly efficient and practical family of
algorithms.

Teaching methods

Teaching method Contact Online
Lectures 28 h 0 h
Exercises 14 h 0 h

Modes of study

Option 1
Available for:
  • Degree Programme Students
  • Other Students
  • Doctoral Students
  • Exchange Students
Written exam
  • In English
Project / practical work
  • In English
English:
Written exam and Project / practical work

Written exam, exercises, and Project / practical work.

Evaluation

Numeric 0-5.

Study materials

Yaser S. Abu-Mostafa, Malik Magdon-Ismail, Hsuan-Tien Lin

Learning from data

http://amlbook.com

Video lectures from the Caltech learning from data online course

Belongs to following study modules

Department of Future Technologies
Department of Future Technologies
2013–2014
Teaching
Archived Teaching Schedule. Please refer to current Teaching Shedule.
Department of Future Technologies
DP in Computer Science
DP in Electr. and Communication Technology
Finnish Study Modules