Optimization of Cosmological Simulations with Artificial Intelligence

Galaxy collisions are an important part of the large-scale structure of the Universe and an important catalyst for intragalactic processes like star formation. Therefore, realistic models of these interactions are an important part of any theory that plans to accurately describe the evolutionary processes of the Universe and, given the size of the problem, efficient computation and data analysis are key. This dissertation presents a proof-of-concept that an artificial intelligence suite, nominally composed of a genetic algorithm and neural network, can optimize the search of the galaxy collision parameter space. Furthermore, this dissertation discusses the possibility that this method can be used for any large problem dependent on a large number of tunable parameters.