8000 GitHub - mehmetbalci/gameAI: Game AI lecture at B-IT [Uni Bonn (2018)]
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content

mehmetbalci/gameAI

 
 

Repository files navigation

GameAI

What is this repository for?

This project is used to manage all code developed for GameAI lecture at Uni Bonn.

Version: 0.2 (second project)

Project 01

Check out the project description.

Task 1

  • Understand tic-tac-toe.py
  • Realize statistical 'strategy' for player X:
  • Collect cell participation for wins
  • Normalize count data
  • Realize heuristical 'strategy' for player X:
  • Evaluate move (Can I win? -> Can enemy win? -> Best probability)

Task 2

  • Realize Connect4 similar to tic-tac-toe.py
  • Realize GUI with user input
  • Try to collect 'good move' statistics

Project 02

Check out the Project description.

Task 1

  • Compute number of nodes in TicTacToe game tree
  • Explain difference in numbers
  • Compute complete game tree ignoring symmetries
  • What is the number of nodes?
  • How many nodes correspond to a win for X?
  • What is the branching factor of the tree?

Task 2

  • Implement MinMax algorithm
  • Use the implementation to compute some MinMax results
  • Consider tree with given MinMax value to its nodes
  • Discuss a checking for alternatives in cases of ties

Task 3

  • Implement MinMax search for Connect4
  • Create an evaluation function
  • Let the computer play a tournament and collect statistics
  • Bonus: Update connect_four.py to implement any size of field

Task 4

  • Implement controller for breakout
  • Modify breakout.py to increase speed over time

Task 5

  • Implement a program to read map data from a .txt file
  • Put map data into a graph
  • Implement Dijkstra's algorithm
  • Implement A* algorithm and use Euclidean distance

Project 03

TBD

Presentation

The presentations are held using Jupyter Notebook. To start the presentation just run:

jupyter nbconvert Presentation.ipynb --to slides --post serve

How do I get set up?

To get the project on your system:

  • You need to have Python 2.7.x or Python 3.x
  • You need git to be installed on your system
  • Run git clone https://bitbucket.org/LeKono/game_ai.git to clone the repo from BitBucket
  • Run git clone https://github.com/LeKono/gameAI.git to clone the repo from GitHub

Get it started:

  • Run pip install -r requirements.txt to install all python modules required to run the code.

  • python tic_tac_toe_B.py starts a game of TicTacToe on terminal

  • from tic_tac_toe_B import TicTacToe is how you can import TicTacToe to a python environment.

  • python connect_four.py starts a GUI

  • from connect_four import ConnectFour can be used to import ConnectFour to a python environment.

The Team

About

Game AI lecture at B-IT [Uni Bonn (2018)]

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published

Languages

  • Python 100.0%
0