Archive for July, 2008

Progress: Graph Comparisons

Posted by Sam on July 30th, 2008 under Honours Project Progress  •  No Comments

I need to write down some stuff here before I lose it.
Currently implementing a menu for dealing with performance comparisons by showing on the graph how the performances stack up against one-another. The menu includes saving current performance (the total performance), loading existing performances (shown in a different colour on the graph when viewing total [...]

Progress: GUI V1.0 nearly complete

Posted by Sam on July 29th, 2008 under Honours Project Progress  •  No Comments

Well, it’s been a while and my post rate has certainly slowed down. But alas, I have been busy working. Just not informing the internet about it.
I have a GUI up and running with some bugs in it. However, the main part of it works and it output’s a performance graph, so it already does [...]

Progress: Self Consolation

Posted by Sam on July 19th, 2008 under Honours Project Progress  •  No Comments

After looking around at my competitors details and personal websites, I realised that my competition was fierce. So, partly giving myself some credit for coming 5th against these guys, and partly finding stuff to put into my Final Report, I am listing my competitor details.

Loria INRIA - MAIA

Bruno Scherrer - Senior researcher, written many publications
Christophe [...]

Progress: Final Results!

Posted by Sam on July 15th, 2008 under Honours Project Progress  •  No Comments

Well, the competition is over and according to the graph, I came in 5th. Out of 15 competitors. That’s ok, but I would have preferred better. Still, you take what you get and be grateful for it. I’ve learnt a whole lot about RL and Tetris (more than one should know), and coming 5th in [...]

Progress: Rather Late Post

Posted by Sam on July 6th, 2008 under Honours Project Progress  •  No Comments

Well, after the previous post, I did write a (or part of a) data file, and ran that through WEKA. The results given didn’t help at all (likely due to the data representation) so I just ran another proving run with the last working agent and after that, ran my testing run.
Unless one of the [...]