cec2002gxk.doc - An Investigation using Co-Evolution to...

Info icon This preview shows pages 1–2. Sign up to view the full content.

View Full Document Right Arrow Icon
An Investigation, using Co-Evolution, to Evolve an Awari Player James Edward Davis and Graham Kendall Automated Scheduling, Optimisation and Planning Research Group School of Computer Science and IT, Jubilee Campus, University of Nottingham, Nottingham, NG8 1BB, UK [email protected] http://www.cs.nott.ac.uk/~gxk Abstract – Awari is a two-player game of perfect information, played using 12 “pits” and 48 seeds or stones. The aim is for one player to capture more than half the seeds. In this work we show how an awari player can be evolved using a co-evolutionary approach where computer players play against one another, with the strongest players surviving and being mutated using an evolutionary strategy (ES). The players are represented using a simple evaluation function, representing the current game state, with each term of the function having a weight which is evolved using the ES. The output of the evaluation function is used in a mini-max search. We play the best evolved player against one of the strongest shareware programs (Awale) and are able to defeat the program at three of its four levels of play. 1. Introduction Game playing has a long history within AI research. Chess has received particular interest culminating in Deep Blue beating Kasparov in May 1997, albeit with specialised hardware [ 1 ] and brute force search, which managed to search up to 200 million positions per second. However, chess is still receiving research interest as scientists turn to learning techniques that allow a computer to ‘learn’ how to play chess, rather than being ‘told’ how it should play [ 2 ]. Learning techniques were being used for checkers (draughts) as far back as the 1950’s with Samuel’s seminal work ([ 3 ], re- produced in [ 4 ]). This would lead to Jonathan Schaeffer developing Chinook, which won the world checkers title in 1994 [ 5 ]. Like Deep Blue the question of whether Chinook used AI techniques or not is open to debate. Chinook had an opening and end game database and in certain games it was able to play the entire game from these two databases. If this could not be achieved, a form of mini-max search, with alpha-beta pruning was used. Despite Chinook becoming the world champion, the search has continued for a checkers player that is built using “true” AI techniques. Chellapilla and Fogel ([ 6 ],[ 7 ],[ 8 ]) developed Anaconda, so named, due to the strangle hold it places on its opponent. It is also called Blondie24 [Error: Reference source not found] which was a name given to the program late in its life in an experiment to see if the name affected the types of player it would attract when playing over the internet and if the other players would treat it differently to a program named something like ‘David0203’. Anaconda (Blondie24) uses an artificial neural network (ANN), with 5046 weights, which are evolved via an evolutionary strategy. The inputs to the ANN are the current board state, presented in a variety of spatial forms.
Image of page 1

Info iconThis preview has intentionally blurred sections. Sign up to view the full version.

View Full Document Right Arrow Icon
Image of page 2
This is the end of the preview. Sign up to access the rest of the document.

{[ snackBarMessage ]}

What students are saying

  • Left Quote Icon

    As a current student on this bumpy collegiate pathway, I stumbled upon Course Hero, where I can find study resources for nearly all my courses, get online help from tutors 24/7, and even share my old projects, papers, and lecture notes with other students.

    Student Picture

    Kiran Temple University Fox School of Business ‘17, Course Hero Intern

  • Left Quote Icon

    I cannot even describe how much Course Hero helped me this summer. It’s truly become something I can always rely on and help me. In the end, I was not only able to survive summer classes, but I was able to thrive thanks to Course Hero.

    Student Picture

    Dana University of Pennsylvania ‘17, Course Hero Intern

  • Left Quote Icon

    The ability to access any university’s resources through Course Hero proved invaluable in my case. I was behind on Tulane coursework and actually used UCLA’s materials to help me move forward and get everything together on time.

    Student Picture

    Jill Tulane University ‘16, Course Hero Intern