0

I am currently trying to train my backpropagation to classify 150k training pairs. Each training pair is a vector of 18 Bipolar numbers and it runs through 2 hidden layers with a final output of 1 number (18-18-18-1).

When I feed my neural network with only a couple thousand training pairs it can classify them 100% but when I try to feed more it makes many errors. Does anyone have any recommendations for what can help a back prop net handle more training pairs? Should I do batch updating of weights or update them for every training pair (what I do now)? Also whats the best way to test for convergence with so many training pairs (when summing up error it's ~40k). Any help is much appreciated.

FYI: each training pair is a loan account I am trying to evaluate with data on balance, age, how long it's been open, average income of zipcode and the target is whether or not it was paid.

Shaido
  • 27,497
  • 23
  • 70
  • 73
matt standley
  • 11
  • 2
  • 4
  • Getting 100% accuracy is overfitting. The NN has not generalized the learning. If a new sample is given, it will not classify properly. Are you getting 100% accuracy for any couple of thousand input pairs or only for a particular set. Is it a real time data? I doubt, your training sample is not properly sampled. – narasimman May 12 '17 at 02:11
  • I only get 100% for the training pair set not any particular set. How would my training sample not be properly sampled? I am grabbing inputs that are not similar to each other. – matt standley May 12 '17 at 02:16
  • if that is not the case, then try batch updating. You can also use dropouts, i.e. severe some of the connections in the hidden layers (making the weights to zero randomly, say 10% of the weights. – narasimman May 12 '17 at 02:18

0 Answers0