GSoC sachith500 Proposal AI

From The Battle for Wesnoth Wiki
Revision as of 01:47, 2 March 2014 by Sachith500 (talk | contribs) (Robotgame AI game)


This page is related to Summer of Code 2014
See the list of Summer of Code 2014 Ideas



This is a Summer of Code 2014 student page


Description

Sachith Seneviratne - AI Global Strategy Proposal

Currently the AI has no way of making an overall decision on whether to attack or retreat. This project will attempt to implement a solution to this problem, by using evaluating functions (or a combination thereof) to provide the AI with the capability of making such a decision.

About Me

Hello everyone! I'm a 22 year old guy studying Computer Science and Engineering at the University of Moratuwa, Sri Lanka. This is my first year taking part in GSoC and I definitely intend to take part in it next year as well. If you wish to contact me I hope the following details will be sufficient. Please feel free to leave me any feedback regarding this proposal, or anything else! :D

IRC

sachith500

Email

sachith500@gmail.com

Github

https://github.com/sachith500

Experience

AI Experience

I have a lot of experience working on AI programs, both swarm-oriented as well as single unit-oriented. Here I will list mostly links to these previous projects (most of which I did out of my own interest).

Google AI Challenge

sachith500's profile: Google AI Challenge 2010
language : Java

Robotgame AI game

language : Python

Kindybot

  • Kindybot
    • Attempts mild aggression, with some risk-taking.
    • Current ELO: 1902
    • Current Ranking: 63

Juggerbot

  • Juggerbot
    • High aggression, less concerned about saving bots.
    • Current ELO: 1884
    • Current Ranking: 66

Wimpybot

  • Wimpybot
    • very low aggression, attempts to flee whenver possible.
    • Current ELO: 1938
    • Current Ranking: 61

Machine Learning Android App

Technologies used: Android Awards: Most Creative Award at Codefest-2013

DiscoverLanka

A location based, content aware Android application with a machine learning algorithm which incorporates an ontology database to provide customized recommendations for tourists. The core learning happens at two places, regarding locations as well as people. Query mining (searches, as well as clicks) is used as the input point of new data to the system. Due to the large amount of input for places (many users searching for same place), the location database is expected to equilibriate first, and this in turn provides additional accuracy to the learning process with regards to users.