Why Neural Network Output is different every time???

In Neural network tool Why it shows every time with different output???

eg:plotsomhits

Answer 1

[Greg Heath]

Initial weights and data division are random unless you initialize the random number generator to the same state before training.

[M J]

Hi, would using something like :

rng(0,'combRecursive');

right before training the CNN work ?

Thanks a lot !

[Greg Heath]

I just use

close all, clear all, rng(0)

as the 1st line of my code.

Hope this helps

Greg

[M J]

Thanks a lot for the answer. The thing is that I intend to randomly shuffle my data as part of the training process, and compare validation accuracies associated with different training/validation combinations. Wouldn't that (writing it as the 1st line of the code) affect the random shuffling part too?

Best.

Answer 2

[SIDDHARAM GADDI]

just simply paste this code before traning the data..

rng('default');

You will get same result every time....

[Carlos Vargas]

Thank you so much, friend! It works so perfect!