A NEW CONCEPT FOR TEACHING AI USING AS EXAMPLE CLASSICS FROM ELECTRONIC GAMES

摘要

There is a problem related to the introduction of Artificial Intelligence in computing courses, where often the student cannot fully absorb its functioning and object of study. Therefore, this paper proposes a practical tool to demonstrate the operation of Artificial Intelligence, showing the machine learning through data mining and generation of decision trees applied in electronic game simulations that were inspired by the most established game styles (like platform, puzzle and racing games) in order to facilitate its understanding by the results obtained. For this purpose, adverse situations were built in which the characters of the games would be submitted.