understanding output from opencv neural network

Asked Dec 11 '13 at 20:10

Active Dec 11 '13 at 20:10

Viewed 290 times

I am testing the opencv neural network. The training input I am using is

Finding the mean of the two numbers and the third one is the expected output. I am using 3 levels with 2 neurons for the input, 3 for hidden and 1 for output with equal weights. I am using back propagation technique.

The output I am getting for

Predict(0,0);  ----> 13.121479
Predict(40,60);  ----> 6.043900
Predict(120,180);  ----> 5.801756
Predict(10,5);  ----> 10.095689

can someone explain the prediction results I am getting.

asked Dec 11 '13 at 20:10

user973743

can you give us the weights of all neuron-connections? – Micka Dec 12 '13 at 10:11
http://stackoverflow.com/questions/1565115/approximating-function-with-neural-network – Micka Dec 12 '13 at 10:37
weights are all same and i set them to 1 – user973743 Dec 12 '13 at 17:22

understanding output from opencv neural network

0 Answers0