Hochschule Darmstadt - Fb Informatik

Drucken| Layout| Design| Schriftgröße English|
Modulbeschreibung
Module:Visual Pattern Recognition
Module numbers:41.5044 [PVL 41.5045; Module 41.50440]
Language:english
Study programme:Dualer Master 2013 - Katalog AS: Anwendungs- und systemorientierte Module
Master 2013 - Katalog AS: Anwendungs- und systemorientierte Module
MN Data Science 2016 - Katalog DS-I: Data Science - Informatik
Type of course:V+P = Lecture+Practical
Weekly hours:2+2
Credit Points:6
Exam:written exam (writing examination)
Registering for examexplicitly and independent of booking
PVL (e.g. Practical):graded (Software demonstration, documentation and presentation)
PVL percentage:50%
Frequency of offering:each winter semester (not yet offered)
Required knowledge:Basic knowledge of graphical data processing
Learning objectives:
  • Students can conceptualise systems for applied pattern recognition problems, especially in the area of medical decision support.
  • Students can conceptualise and implement feature extraction techniques to represent colour, shape, and texture properties of objects in images.
  • Students can conceptualise and implement supervised classification algorithms (e.g., linear classification, Support Vector Machines, Bayes classification, deep learning, etc.)
  • Students can perform the training phase in supervised classification tasks including data selection, acquisition, and labelling.
  • Students can conceptualise and implement unsupervised pattern recognition algorithms, especially sequential, fuzzy, and hierarchical clustering schemes.
  • Students can plan, perform, and interpret experiments for comparative evaluation of pattern recognition systems.
Content:
  • Feature extraction, selection, and transformation (e.g., Fourier and wavelet transform, Principal Component Analysis, etc.)
  • Supervised pattern recognition algorithms including linear classification, Support Vector Machines, and Bayes decision scheme.
  • Unsupervised pattern recognition techniques including sequential, hierarchical, and fuzzy clustering.
  • Adaptivity in pattern recognition methods including the concept of relevance feedback.
  • Evaluation of pattern recognition systems
Literature:
  • Theodoridis and Koutroumbas, Pattern Recognition, Elsevier, 2009, ISBN: 978-1-59749-272-0
  • Grzegorzek, Sensor Data Understanding, Logos, 2017, ISBN: 978-3-8325-4633-5
  • Bishop, Pattern Recognition and Machine Learning, Springer, 2006, ISBN: 978-0387-31073-2
  • Goodfellow, Bengio, and Courville, Deep Learning, MIT Press, 2016, ISBN: 978-0262-03561-3
Lecture style / Teaching aids:Lecture
Practical projects in small groups of 2 to 3 students
Materials: slides, papers, demonstrators
Responsibility:Marcin Grzegorzek
Released:WS 2018/2019
Professional competencies:
  • formal, algorithmic, mathematical competencies: high
  • analytical, design and implementation competencies: medium
  • technological competencies: medium
  • capability for scientific work: high
Interdisciplinary competencies:
  • project related competencies: medium
  • interdisciplinary expertise: basic technical and natural scientific competence, basic economic competence
  • social and self-competencies: ability to work in a team, analytical competence, judging competence, deciding competence

[Fachbereich Informatik] [Hochschule Darmstadt]
© 2008 - 2022 FBI OBS Team. Alle Rechte vorbehalten.