Use of Genetic Algorithms in Pattern Recognition Studies

Use of Genetic Algorithms in Pattern Recognition Studies

02 OCTOBER 2018 (Morning)


Dr. Yalçın İŞLER

Duration: 4 hours


Contents: To find out the hidden information (pattern) from a raw data is called as “Pattern Recognition”. A general pattern recognition system consists of six stages: data acquisition, preprocessing, feature extraction, feature selection, classification and evaluation. A problem-specific pattern recognition system is achieved when an optimal combination of parameters and classification model in these stages. “Genetic Algorithms (GA)”, which is one of stochastic methods, is an optimization algorithm. GA imitates the basic principles of natural evaluation and becomes one of commonly used algorithms in the literature. GA can work with noisy and even non-derivative big data clusters. It is also capable to work in parallel and distributed computational systems. In this workshop, a theoretical introduction to GA with following pattern recognition applications will be given in Matlab programming environment.

- Finding minimum (or maximum) values of a function using GA

- Determining coefficients in linear regression using GA

- Determining coefficients of an arbitrary function using GA

- Sample size reduction using GA

- Feature selection using GA

- Training of perceptron algorithm using GA

- Training of Multi-Layer Perceptron Neural Networks using GA


Prerequisites: All attendees should bring their own notebooks (fully-charged and even charging apparatus) and install Matlab version 2017a before the workshop.


Special Note: You can take photo and video freely if you share original files by Yalçın İŞLER, and you can publish in social media if you tag him, of course :).