Share Site Map Contact us Home Page
Home > The Learning Machine Challenge > LMC2 Quick Overview


LMC2 Quick Overview

LMC2 General Information

LMC2 Official Rules

LMC2 SDK Specs

LMC2 Protocol Specs

LMC2 Testing Zone

Enter LMC2

LMC2 FAQ
LMC2 Quick Overview
  Printable version
Ai Discussion Boards  Talk about it


Imagine playing a language game with an unknown opponent. It's your turn to say something. The only thing you know for sure is the symbol set ("alphabet") used in the game, as well as the history of the game until this point.
 
You make a move by "saying something": by sending one or more of the alphabet symbols you know about to the outside world (represented by your opponent, also known as "the judge" or "the trainer").
 
You receive a score from the judge/trainer, representing the quality of your response. The score is one of a discrete set of scores between -1 and 1.
 
You may also receive some guidance from the judge, informing you of the expected response - the one that would have entitled you to the highest score for your last response.
 
Do you think you could learn to play this game well? Even if you don't know anything about the judge and the language he is using? The Learning Machine Challenge is a quest to find the best learning computer programs which are capable of doing just that.
 
Now that you have been introduced to the Learning Machine Challenge, you might want to read some more general information about it.
 
If you have any questions about the Learning Machine Challenge, post them on the Ai Forums.