Skip to content

Latest commit

 

History

History
43 lines (38 loc) · 1.3 KB

readme.md

File metadata and controls

43 lines (38 loc) · 1.3 KB
Raw

Intelligent Agents for the World of Wumpus

About

This project is a work proposed in the computational intelligence discipline. Developed by computer engineering students at the Federal University of Pará - Tucuruí.

Students:

@ Denys Menfredy @ Jeremias Kalebe @ Renuá Meireles @ Rodrigo Moraes

Objectives

  • The main goal is to develop intelligent agents to act in the environment of World of Wumpus
  • Use a Genetic Algorithm (GA) as learning mechanism.

Enviroment

  • NxN Size;
  • N-1 Pits;
  • 1 Wumpus;
  • 1 Gold;

Enviroment

  • Actions:
    Code Meaning
    N Move to North
    S Move to South
    L Move to East
    O Move to West
    P Pick up the Gold
    X Shoot East
    Y Shoot West
    Z Shoot North
    C Shoot South
  • Goal: Seek for gold, and return to the initial coordinate.

Tasks

[x] Develop the Wumpus Environment. [x] Develop an Agent with GA as learning mechanism. [x] Implements an Interface for a better understanding of learning process. [x] Collect data from agent's efficience. [x] Implements Simple Data Visualization.

License

MIT - http://www.opensource.org/licenses/mit-license.php