[R] problem with nnet

From: madhurima bhattacharjee <madhurima_b_at_persistent.co.in>
Date: Fri 03 Feb 2006 - 03:33:50 EST

Hello All,

I am working with samr and nnet packages. I am following the steps given below:
1> I take a input file with signal values for 9506 genes and 36 chips , belonging to two classes.
2> I perform samr analysis on 80% of chip data from both the classes.(selected by random sampling)
3> I then use the data of only the significant genes from this samr analysis to train nnet.
4> The parameters I am currently using for nnet are: result <- nnet(traindata[sortsamp,], targets[sortsamp,], size = nnetsize , rang =0.00000003 ,decay = 0.00009, maxit = 100, MaxNWts =100000)

           traindata is the significant gene's data, sortsamp is the randomly sampled number out of those genes and targets is the class indicator of the significant genes.
5> Then I am using the data from the chips left out from samr(20% chip data) to test the nnet.

      I use the following command for this:
      pred <-predict(result, testdata) #result is the nnet output(given 
above) and testdata is the chipdata of remainder chips. 6>I run the nnet part for 100 runs using the same significant genes in all runs.

The problem is that the pred in each run gives same values for both the classes.
[1] "pred----->"

           cer noncer
[1,] 0.4990032 0.5009930
[2,] 0.4990032 0.5009930
[3,] 0.4990030 0.5009933
[4,] 0.4989994 0.5009968
[5,] 0.4990030 0.5009932
[6,] 0.4990032 0.5009930
[7,] 0.4990032 0.5009930

This is really weird result.

I have checked the data and code several times but couldn't figure out the problem.
I am really stuck in this.
Can anyone help me as soon as possible?

Thanks in advance,

R-help@stat.math.ethz.ch mailing list
https://stat.ethz.ch/mailman/listinfo/r-help PLEASE do read the posting guide! http://www.R-project.org/posting-guide.html Received on Fri Feb 03 03:39:43 2006

This archive was generated by hypermail 2.1.8 : Fri 03 Feb 2006 - 15:12:01 EST