Titolo del progetto di ricerca: "SemInt - Semantic Enrichment and Interoperability"
Descrizione dell'attività da svolgere dal/la collaboratore/trice:
Study, development, and evaluation of an experimental artificial intelligence system to learn best strategies to play a card game. The technology to be used is AlphaGo Zero, a learning algorithm which knows only the rules of the game (of different types to be integrated), and a Montecarlo search method to evaluate the quality of the evolving strategies. The final system will be provided with a graphical user interface to manage the games to be played against the system. There will be an evaluation of the semantics and the quality learned by the system.
The work consists in the analysis, the implementation, the training, and the evaluation of a software based on AlphaZero, which is able to learn the rules and the strategies of a card game. These tasks are organised in autonomy by the collaborator, also with reference to working times and places
Responsabile: prof. Enrico Franconi