Høst 2024

FYS-3012 Pattern recognition - 10 stp

The course is administrated by

Institutt for fysikk og teknologi

Type of course

The course is available as a singular course. The course is also available to exchange students and Fulbright students.

Course overlap

FYS-8012 Pattern recognition 8 ects

Course contents

The course covers data analysis techniques such as Bayes classifiers, estimation of probability density functions and related non-parametric classification approaches. Further, linear classifiers using least squares are addressed, in addition to simple processing units (neurons) and their extension to artificial neural networks. Linear and nonlinear (using kernel functions) support vector machine classifiers are discussed, in addition to feature extraction and data transformation using eigenvector-based methods such as Fisher discriminants. Methods for grouping, or clustering, data are treated in detail, including hierarchical clustering and k-means. Exercises and problem solving, in addition to practical pattern recognition for data analysis using programming, are strongly emphasized. Basic programming skills are required.

Admission requirements

Admission requirements are a Bachelor's degree in physics or similar education, including specialization in physics worth the equivalent of not less than 80 ECTS credits. Local admission, application code 9371 - singular courses at Master's level.

Objective of the course

Knowledge - The student can

describe the concepts of classification, clustering and dimensionality reduction in data analysis
apply advanced methods from pattern recognition to diverse classification, dimensionality reduction and clustering problems
compare different algorithms with respect to their strengths and applicability
describe important pattern recognition applications in society

Skills - The student can

analyse modern pattern recognition methods and apply them in independent practical and theoretical problem solving
train and validate a pattern recognition algorithm for a defined task using available data
analyse Bayes classifiers in terms of error probabilities
design linear classifiers for minimization of squared errors and other criteria
design and analyse nonlinear classifiers in the form of neural networks
perform feature extraction and data transformation, e.g. using eigenvectors
explain different clustering algorithms, and analyse their strengths
implement in practice all methods discussed in the course for analysis of data

General competence - The student can

report the comparison of several pattern recognition algorithms applied to a practical problem
appreciate the importance of pattern recognition in society
work with pattern recognition methods for analysis of real data

Language of instruction

The language of instruction is English and all of the syllabus material is in English. Examination questions will be given in English, but may be answered either in English or a Scandinavian language.

Teaching methods

Lectures: 40 hours Exercises: 40 hours