Deep Learning Techniques Applied to Videogames

Abstract

This document is a reflection of the work done as a final degree project. In summary, the work has consisted of research on techniques capable of generating and simulating artificial intelligence in general, and in particular, learning to use deep learning techniques through the use of neural networks. The project has been divided into three different parts: research about artificial intelligence and its techniques, testing of existing examples in order to learn how to develop related techniques, and application of the previous concepts to a specific topic related to videogames using the previous techniques. Thus, throughout the report each part of the process will be explained in detail following the above order and emphasizing the most important points, starting with the technical explanation of what a neural network is and what it is used for; following with the different examples that have been followed to understand and make use of them; and ending with the development process of a deep learning project capable of estimating the total duration of a game in order to evaluate whether the pairing of the players that form it has been correct or not.