No description available.

No description available.

NeuroEvolution besides deep learning is considered the most promising method to train and optimize neural networks. Neuroevolution uses genetic algorithms to train the controller of an agent performing various tasks. Traditionally, the controller of an agent will be encoded in a genome which will be directly translated into the neural network of the controller. All weights and the connections will be described by their elements in the genome of the agent. Direct Encoding – states if there is a single change in the genome it directly affects a change in the brain. Over time, different forms of encoding have been developed, such as Indirect and Developmental Encodings. This paper mainly concentrates on Developmental Encoding and how it could improve NeuroEvolution. The No-Free Lunch theorem states that there is no specific optimization method that would outperform any other. This does not mean that the genetic encodings could not outperform other methods on specific neuroevolutionary tasks. However, we do not know what tasks this might be. Thus here a range of different tasks is tested using different encodings. The hope is to find in which task domains developmental encodings perform best.

The aim of the thesis is to verify synergy of genetic programming and neural networks. Solution is provided by set of experiments with implemented library built upon benchmark tasks. I've done experiments with directly and also indirectly encoded neural netwrok. I focused on finding robust solutions and the best calculation of configurations, overfitting detection and advanced stimulations of solution with fitness function. Generally better solutions were found using lower values of parameters n_c and n_r. These solutions tended less to be overfitted. I was able to evolve neurocontroller eliminating oscilations in pole balancing problem. In cancer detection problem, precision of provided solution was over 98%, which overcame compared techniques. I succeeded also in designing of maze model, where agent was able to perform multistep tasks.