Home > Search Results: keywords:"neuroevoluce|umělá inteligence"
Search:
Search Tips :: Advanced Search
Search collections:
Sort by: Display results: Output format:
National Repository of Grey Literature 1 records found  Search took 0.01 seconds. 
Neuroevolution-based AI for the Dominion game
 Machala, Patrik ; Kuboň, David (advisor) ; Holan, Tomáš (referee)
The subject of this thesis is a simple user interface for the base version of the card game Dominon and the development of an artificial intelligence capable of playing this game. The AI is designed regardless of the initial configuration of the game. That allows an immediate start without waiting for the evolution of the opponent. The basis of this scheme is a recurrent neural network evolved by neuroevolution of its weights. Its inputs are made from representation of current game state and its output is the valuation of cards, which leads to their purchase. Second part of a player's move, the so-called action phase, is controlled by heuristics. The thesis is not limited to a single method of AI development but compares different types of evolution and different numbers of neurons in the hidden layer. According to the completed experiments the neural network evolved by competitive co-evolution with populations swapped according to their skill was the strongest opponent. Results of AI development are then compared with conclusions made on the basis of other artificial intelligencies for the Dominion game with most of these confirmed.
Detailed record

Interested in being notified about new results for this query?
Subscribe to the RSS feed.
Digital Repository :: Search :: Submit :: Personalize :: Help
Powered by Invenio v1.1.2
Maintained by admin.invenio@techlib.cz
This site is also available in the following languages:
Česky  English