Machine learning based on logic and probability

Cilj ovog ovog rada je istražiti i pokazati načine primjene strojnog učenja u svrhu modeliranja Texas Hold’em Poker agenta koji će igrati na zadovoljavajućoj razini. U prvom poglavlju rada objašnjen je pojam strojnog učenja zajedno s njegovim granama i metodama rješavanja problema određene prirode. Texas Hold’em Poker je kompleksna igra jednostavnih pravila što je pokazano u drugom poglavlju. Objašnjeni su najvažniji koncepti Texas Hold’em Pokera čije znanje može biti primjenjeno kako bi model igrača bio što efikasniji. Konačno, navedeni su neki od dosadašnjih pokušaja rješavanja ovog problema zajedno s načinom njegovog rješavanja i mjerom uspjeha, tj. rezultatima.The aim of this paper is to explore and demonstrate how to use machine learning in order to simulate/ make a model of a Texas Hold'em Poker agent that will play at a satisfactory level. In the first chapter of the paper, the concept of machine learning is explained, along with its branches and methods of solving different problems. Texas Hold'em Poker is a complex game with simple rules, as it is shown in the second chapter. Furthermore, the most important concepts of Texas Hold'em Poker are explained. The knowledge of these concepts can be applied to make a model of the player more efficient. Finally, some of the past attempts to solve this problem are mentioned together with the way it is solved and the measure of success, i.e. the result