CONSTRUINDO UM JOGADOR AUTÔNOMO: IA de aprendizado em jogos usando JavaScript

Abstract

The present work presents an experience report on the construction of an Artificial Intelligence for learning in games, experienced in the context of technology, demonstrating the machine learning of an autonomous player during the months of February and March 2024. The objective of the research consists of introduce machine learning, which is a field of artificial intelligence that focuses on the development of algorithms and techniques that allow

19

computers to learn from data, using tools that allow you to visualize this learning. Through this narrative, an autonomous player was developed to demonstrate the evolution of machine learning in a game, using tools used in web development. The construction of artificial intelligence was analyzed and described, demonstrating its learning. The experience was theoretically based on teaching “ladybugs” to move forward by avoiding “predators” using frontal sensors. According to Endo and Tanaka (2023), recent studies show that they are testing AI methods in current games, due to to vast uncertainty and complexity. Complexity is a relevant factor in studies of artificial intelligence development algorithms, as the more complex the game, the greater the difficulty of research and the inclusion of artificial intelligence. Furthermore, machine learning has contributed for improvement in various fields of activity. This is the result of advances in recent decades that have allowed innovative projects based on artificial intelligence. It is believed that the results of this report can contribute to the debate and reflections on artificial intelligence and its impact on machine learning in games.