Pattern Recognition

  • Code:  
  • Number of credits:  5
  • Subject coordinator:
    Victor MATROSOV
  • Lecturers: Viktor GORELIK,
    Dmitriy BORODIN
  • Credit contract possible:  Yes
  • Examination contract possible:  Yes
  • Teaching language:  English, Russian

 

 Course Activity: 
  • Lectures
  • Practical Classes
  • Lab Classes

 

 Course unit type: 
  • Specializing

 

 Content:

  1. General description of the objects and events recognition problem
  2. Standard problems for recognition systems development
  3. Recognition systems classification
  4. General definition of the recognition problem
  5. A priori information processing methods
  6. Probabilistic recognition systems
  7. Constructing the vocabulary of features
  8. Logical recognition systems
  9. Recognition algorithms based on the values calculation
  10. Object recognition process management
  11. Recognition systems efficiency

 

 


  Objectives:

 A. General competences 

PB - 01. Intellectual and cognitive capacities
PB - 02. Acquisition and processing of information

Explanation

 PB – 01. The student is able to classify objects and phenomena and use corresponding features to describe them

PB – 02. Depending on the type of a priori and a posteriori information the student is able to choose and use appropriate methods of recognizing real life objects and phenomena

  B. Profession-oriented/ General scientific competences  

Explanation

  • Know basic methods: statistical, logical, algebraical, learning processing, self-learning processes, neural networks, genetic algorithm
 C. Profession-specific competences  
 Linear and nonlinear decision rules, Bayes criterion, Pirson criterion, Wald’s statistical analysis, Boolean algebra methods, simulating modeling.
 D. Scientific competences 

/

Explanation

/

 

 Prerequisites:

 A. Previously required courses 

Mathematical Analysis, Algebra, Theory of Probability and Mathematical Statistics, Discrete Mathematics, Mathematical Logic, Theoretical Basis of Informatics, Geometry, Numerical Methods

  B. Required competences  

Mathematical Analysis: the methods of working with analytical functions, convex analysis, vector analysis, methods of constructing pivotal separating hyper plane. Algebra: the calculation with letters, matrices and special products, solving of equations of the first, second and n-degree, solving of simultaneous equations. The Differential Analysis: finding derivations of the first and second degree and complex derivations, working with integrals of the first degree. Theory of Probability and Statistics: theorem of probability calculation, conditional expectation function, variance, probability distribution of conditional density, the law of large numbers, regression analysis method, statistical criteria.

Calculation methods: branch-and-bound method, set of equations solving methods.

Boolean algebra: predicate calculus.

 

 Educational tools:

 A. Type 
  • Course
  • Textbook
  • online learning platform (dokeos, private lecturer website)
  • Internet
  • mathematical computing systems: Mathcad, Mathlab
  B. Obligatory educational tools  

course book: Methods of Recognition by A. GORELIK, V. SKRIPKIN (both in English and Russian) Mathematical computing systems: Mathcad, Mathlab

 C. Recommended educational tools 
 Textbooks:
  • Ferdinand van der Heijden, Robert P. W. Duin, Dick de Ridder, and David M. J. Tax. Classification, Parameter Estimation and State Estimation: An Engineering Approach Using MATLAB. Wiley, Chichester , UK, 2004.
  • Daniel Peña. Análisis de datos multivariantes. McGraw-Hill, Madrid, February 2002.
  • Alvin C. Rencher. Methods of Multivariate Analysis. Wiley-Interscience, 2 edition, 2002.
  • Richard O. Duda, Peter E. Hart, and David G. Stork. Pattern Classification. Wiley-Interscience, New York, 2 edition, 2001.
  • Tou J.T., Gonzalez R.C., 'Pattern Recognition Principles', Addison Wesley, 1979
  • Schalkoff R.J., 'Pattern recognition: statistical, structural and neural approaches', Wiley, 1992
  • Namboodiri N.K., Carter L.F., Blalock H.M., 'Applied Multivariate Analysis and Experimental Designs.', McGraw-Hill, 1975
 Online learning platform: http://dokeos.mpgu.edu

 


 Teaching methods : 

 A. Type 
  • Lectures
  • Practical Classes
  • Lab Classes
  • Other: higher education on distance
  B. Description  

This course is taught in English or Russian

Plenary lectures: 36h = 2h per week during 18 weeks during the first semester.

Practical Classes (obliged): 18h = 2h per two weeks during 18 weeks during the first semester.

Lab Classes: (obliged): 18h = 2h per 2 weeks during 18 weeks during the first semester.

Higher distance education: for this course, an extra syllabus is available, containing guidelines how to study the material in an optimal way.

 


 Assessment : 

 A. Type 
  • permanent (written tests during semester)
  • oral examination
  • evaluation
  B. Description  

Total number of credits: 5 credits
Period of examination 1 (first chance)
PR: 3 credits - 1 oral assessment
PR labs: 1 credit - permanent and several computing assessments
PR practical classes: 1 credit - permanent and several written assessments
Different approach: 1 oral exam, the mark you get for the exam gives you the corresponding number of credits: 5 credits for A, 4 credits for B and 3 credits for C.
Examination contract:
1 written assessment for the OR. The mark obtained in this examination is the result of the OR and the OR labs.
Higher distance education:
1 written assessment for the OR. The number of credits obtained in this examination is the result of the OR written examination.

 

 Teaching support : 

Extra course OR: 2h per week during 18 weeks during the first semester.
Higher distance education: for this course, an extra syllabus is available, containing guidelines how to study the material in an optimal way.