unable to apply regression classification

Hi,

I am trying to apply regression on given data, In my case the given data is 100x100 matrix in which rows are defined as inputs to my system and coulms are the instances.

The problem is that its is unable to train system, and is not giving output(y) according to desired output. Can anybody help me or guide me I am uploading my code as well

for desired output(check) I use different stratigies

1: In this method I give a threshold value. so in all columns it counts those values which are greater than threshold are sumed and in rest of the coulmn where the values are below threshold are replaced by zero and finally the columns which contain values above threshold are replaced by 1 by using this command.

check=sum(xtrain>1.895999999999986e-09);
check(check<10)=0;
 check(check>=10)=1;

2: secondly I apply multiple threshold limits and gives some random values

for z=1:1:100
 if (model(1,z)<1.380000000000160e-10) 
     check(1,z)= -0.5;
 elseif  (model(1,z)<1.020000000000470e-10)
     check(1,z)= -0.8;
% %  x<1.048000000000030e-09 = 0.2
 elseif  (model(1,z)<1.572000000000080e-09)
     check(1,z)= 0.1;
 elseif (model(1,z)<2.016000000000060e-09)
     check(1,z)= 0.5;
 else  (model(1,z)>2.016000000000060e-09) 
     check(1,z)= 1;
 end
end