A study of cell-based genetic algorithms with applications to neural networks

In the past few decades, Genetic Algorithms (GA) and Artificial Neural Networks (ANN) have been widely used in various fields, such as data processing, robotics and pattern recognition. In particular, GA is used in optimization problems for which there is no known tractable algorithm for finding solutions, while ANN is used mainly in classification or regression problems. To improve the effectiveness of ANN, scientists proposed methods of combining GA and ANN. While most previous approaches apply GA to obtain the structure or the initial weights of an ANN, this thesis explores the cellbased GA model and its application in training an ANN as part of a supervised-learning task.