BrainMeta'   Connectomics'  

Welcome Guest ( Log In | Register )

 
Reply to this topicStart new topic
> Problem with Neural Network
eisbergsalat
post Jan 22, 2007, 01:11 PM
Post #1


Newbie
*

Group: Basic Member
Posts: 1
Joined: Jan 22, 2007
Member No.: 7188



Hello,

first, excuse my engish, please. :-)

I am trying to train a Neural Network to be an opponent to a human player in the game HEX (see wikipedia: Hex).
To do this, i have used the same idea as which was already used for TD-Gammon: I am using a free-forward Net with one Input layer (Boardsize*Boardsize+1), one hidden layer (Boardsize*Boardsize/2.3) and one output layer (only 1 row), all sigmoid layers. The net shoud assign the winning-probability to a board-configuration (i.e. 0.0 if it seems, that player 2 will win the game and 1.1 if it seems, that player 1 will win the game).
The board is presented in the following way: 0.0 if the position is taken by player 2, 1.0 if the position is taken by player 1 and 0.5 if the position is empty.
Actually, with a 8x8 Board the net should do the following:

array of 64+1 double values between 0.0 and 1.0 (+1 for the move) -------> value between 0.0 and 1.0

My question is the following: Have you got a clue, how i can determine such values as 'learning rate', the numbers of hidden units, momentum, etc.? What is about two or three hidden layers?
The number of training patterns will be about 1000 and (second net) about 100 000.

Have you got an (heuristic) idea for the values or any other hint?

Thank you for your help,
florian
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
maximus242
post Jan 23, 2007, 02:24 PM
Post #2


God
******

Group: Basic Member
Posts: 1755
Joined: Jan 24, 2006
Member No.: 4768



Learning rate that is too high turns network to crap, to low and it takes forever. You will probably need at least 10,000 repetitions for training, bare minimum. Umm I think regular learning rate is 0.7? Its been a while since doing NN so if im wrong, im sure lucid will correct me. The size of a neural network determines how much it is able to calculate, generally larger NN are capable of more but are slower, smaller ones are faster but are capable of less. Two or Three hidden layers in neural networks is dependant on the design of the NNet.
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
D.R
post Oct 22, 2007, 03:46 PM
Post #3


Newbie
*

Group: Basic Member
Posts: 12
Joined: Oct 22, 2007
Member No.: 13704



It depends to the model and designer and if the Neural Network its going to learn from the player

Regards.
User is offlineProfile CardPM
Go to the top of the page
+Quote Post
kangyu
post Mar 22, 2010, 10:46 PM
Post #4


Newbie
*

Group: Basic Member
Posts: 1
Joined: Mar 22, 2010
Member No.: 32665



He is clear in what He asks and clear in what He offers. The choice is up to
User is offlineProfile CardPM
Go to the top of the page
+Quote Post

Reply to this topicStart new topic
1 User(s) are reading this topic (1 Guests and 0 Anonymous Users)
0 Members:

 



Lo-Fi Version Time is now: 17th November 2017 - 10:37 PM


Home     |     About     |    Research     |    Forum     |    Feedback  


Copyright BrainMeta. All rights reserved.
Terms of Use  |  Last Modified Tue Jan 17 2006 12:39 am

Consciousness Expansion · Brain Mapping · Neural Circuits · Connectomics  ·  Neuroscience Forum  ·  Brain Maps Blog
 · Connectomics · Connectomics  ·  shawn mikula  ·  shawn mikula  ·  articles