Final Report

[CapnP]

Figured we should get a nice early start on the report as soon as possible.

I volunteer! I'll probably start work on it during the week.

[nelfin]

Cool, thanks Tim. Hopefully we'll have plenty to write about. It shouldn't be too hard to talk for 8 mins either.

[CapnP]

Noting some stuff down to include so I don't lose it:

IF our 3-Stage player (denoted 3J) is easily generalizable to a 92-Stage player (92J) then we should include:

• line graph plotting each feature's weight throughout the game (all on one plot; diff coloured lines for each feature)
  • Epic space filler: multiple plots at a varying number of learning iterations.
  • Not so crazy: only 1 extra plot; only shows 1 feature at a varying number of learning iterations on the same plot
• Win rates between various pairings of (A-n is agent A at depth n)
  • 1J-5, 1J-6, 1J-7
  • 3J-5, 3J-6, 3J-7
  • 92J-5, 92J-6, 92J-7
  • N-7, N-8, N-9 (Straight up naive slutty stone-greedy Negamax. Hopefully will show we can kick ass even with a lower depth)
• Also a runtime analysis of the above pairings

[jpnelson]

That sounds good. I've really got to do my part for this well before the 8th, as my exams start on that day. So I'll spend the next couple of days working on it.