 Test Data with a Trained Network
 

  1. User Inputs:
   a. Testing file name
   b. Weight file for a trained network
   
  2. Testing:
   a. Get the network structure from the weights file.
   b. Using the weights to test the file and calculate the error.
   c. Total MSE and MSE per output are calculated and could be saved. 

  3. Requirment:

   The training file which was used to generate the network structure
   must has same origination, otherwise, the test error could be much
   bigger than training result.

  4. Example: 

  twod.tst            ! Filename to be test
  mlptwod.wts         ! Weights file name


      The program will test the testing file twod.tst suing the trained network
      with weights file mlptwod.wts and result could be saved.





