![]() ![]() What it is being computed is the value of a state, s, at the (k+1)th time step. The interpretation of the above equation is a lot simpler than it looks. The solution below will tackle both of these goals without running a single search algorithm. Pac-man must over time, learn how to maximise the points that can be earned while alive.Pac-man is primarily motivated to win the game by not being eaten by Inky or Clyde.Although this does make it easier for Pac-man to win the game, the objective is two-fold: Inky and Clyde, in this version of Pac-man, are not particularly bright and traverse the game board with random behavior. Eating either ghost after devouring a Power Pellet, results in a reward of 200.Similarly, a death penalty is served to de-incentivise the agent to exit the game by being eaten by the ghost.The considerable increase incentivises the agent to finish the game. If the AI agent is able to complete the game, then there will be an increase of 500. ![]() This is specifically in place to prevent the AI agent from finding itself in a pattern that forever evades Inky and Clyde thereby making it impossible to complete the game. For every change of state, there incurs a living penalty of 1.For every Pellet that is devoured, the score increases by 10.Furthermore, like any good game, there exists a point system that is altered slightly from the traditional game. The actions available to either Pac-man or the ghosts are movement-based actions that dictate whether the entity moves up, down, left, or right. Each state represents the world at a specific point in time and any action performed by either Pac-man or the ghosts will result in a change of state. The interactions between these characters and the world around them are encapsulated within a complete board representation of the game - otherwise known as the state. Figure 1: Pac-man game board (Image by Author). ![]()
0 Comments
Leave a Reply. |
AuthorWrite something about yourself. No need to be fancy, just an overview. ArchivesCategories |