A-Star Search: A customizable implementation of the A-Star search algorithm-the AStar class. It can be applied to any concrete implementation of the State class that can be easily be extended to various problems of AI. Following problems have been solved using this generic implementation (click for an applet demo)-
Tic-Tac-Toe: The TIC-TAC-TOE game with a GUI in the form of an applet, uses a set of rules to determine its next move.
See applet. The code involves 3 classes- a Board class that encapsulates the board and provides one very useful operator winningPos, the TicTacToe class that contains the play method which is the heart of this program and the TicTacToeApplet that provides a simple UI in the form of buttons and a logger. Tic-Tac-Toe-Will-Never-Lose-Strategy.
Traveling Salesperson Problem: A genetic algorithm with user defined parameters is used to search the huge TSP search space and produce a "near-optimal" solution. Do check this
applet that also has a random problem generator and a
and a TSP graph to display the fittest individuals of different generations.
Monkey & Banana Problem: A custom implementation of a bare-bones FOPL inference engine that performs a goal oriented exploration of the set of well formed forumulae and solves the
monkey-banana mystery.