A Reinforcement Learning Approach for Solving KRK Chess Endgames
Reinforcement learning approach used to train a chess agent in order to make him able to play a particular type of chess end game called the KRK one. In this type of position, only 3 pieces are left on the board: the White King, the White Rook and the Black King and the main purpose for the white player is to checkmate its opponent by coordinating the movement of its 2 pieces.
This project is about simulating a method called Deep drawing, which is used to form an object from a single sheet of metal. Deep drawing creates metal shapes by applying force from a punch to a sheet of metal.The punch will move down until the metal is shaped into the required shape.