Ing. Tarin Gamberini

Menu

• Home
• Curriculum Vitae
• State Exam
• University Studies
• Java
• Site's Map
• Email
• PGP/GPG key

Citazione

Counter

Languages

• Italiano
• English

Standards

Campagne

Barra di navigazione

Home > University Studies > University Projects

Artificial Intelligence

Algorithm of the Tagli Alfa-Beta

During the Artificial Intelligence course we have studied a particular set of games which have the following characteristics:

• There are only two players which alternatively execute a move at a time.
• They are perfect knowledge games: the players have both the same informations.

Typical perfect knowledge games are draughts, chess, etc... while typical imperfect knowledge games are the cards one like poker, bridge, etc...

It's possible to describe a game sequence by a tree called states' space, in which every node represents a game state. It starts from an initial state in which is established the player that has the right to move. The player performs a move reaching the successive state in which the right to move passes to the other player. So on till the reaching of the final state: a leaf of the states' space tree in which isn't possible any other move.

A typical approach to solve a game consist in a search in the states' space tree performed by algorithms. During the course we have studied the MIN-MAX algorithm and the Tagli Alfa-Beta one.

In this report we have implemented the Tagli Alfa-Beta in Java language.

The verify of the correctness of the implementation has be performed on exercises of some exam sessions complete with solution.

• Introduzione
• 1 Principio di funzionamento
• 2 Diagramma di flusso dell'algoritmo
• 3 Analisi delle classi
  • 3.1 Descrittore dello stato di gioco
    • 3.1.1 Attributi
    • 3.1.2 Metodi
  • 3.2 Tagli Alfa-Beta
    • 3.2.1 Attributi
    • 3.2.2 Metodi
  • 3.3 Gioco generico
    • 3.3.1 Attributi
    • 3.3.2 Metodi
• 4 Verifica dell'implementazione in Java
• Conclusioni
• A Codice sorgente in linguaggio Java