General Information

The first date for the lecture is 

Tuesday October 17, 2017 in room 0.68 (Seminarraum Werkstoffwissenschaften), 12:15.

The exercises will start on Monday, October 23, more information is given in the first lecture.

Contents 

The lecture will discuss the following topics: 

• the key concept of pattern recognition
• sampling and quantization
• histogram equalization and thresholding
• noise suppression
• edge detection
• non-linear filtering
• normalization
• heuristic feature extraction: Fourier transform, Walsh-Hadamard transform, Haar transform, linear predictive coding, moments, wavelets
• analytic feature extraction: principal component analysis, linear discriminant analysis, optimal feature transform
• optimization: gradient descent, coordinate descent
• feature selection
• introduction to statistical classification
• selection of machine classifiers

The methods and procedures that are presented in this lecture are illustrated in the exercise courses. The tutorials put emphasis on the practical realization of these methods. 

Slides 

The slides of the lecture are available here.

It is absolutely recommended to attend the lectures in order to make personal notes. In addition, covered topics should be completed with personal research in the recommended technical literature and publications that will be announced in the lecture.

Exercises

The web page of the exercise courses is here.

It is absolutely recommended to attend the exercise courses and to do the assigned homework in order to get a deeper understanding of the topics discussed in the lecture.