2

I'm completely a noob in auto-encoder and deep learning thing. And now I'm trying to implement a simpe auto-encoder.

What I want to ask is: If given same dataset, will auto encoder produce the same optimum weight?

Thanks for your help

MCRmy
  • 21
  • 1
    It depends upon your weight initialization and your training algorithm. If you use the same initial weights each time, use a deterministic training algorithm, and do not change the order in which your training data is fed into your algorithm, then I believe you should come up with the same results. – John Yetter Dec 09 '15 at 02:05
  • 1
    @John Yetter I use L-BFGS for training. If I randomize the initial weight, then it's normal if i always get different weight, right? Thanks for your response – MCRmy Dec 09 '15 at 02:16

1 Answers1

1

Partially answered in comments:

It depends upon your weight initialization and your training algorithm. If you use the same initial weights each time, use a deterministic training algorithm, and do not change the order in which your training data is fed into your algorithm, then I believe you should come up with the same results. – John Yetter

( I use L-BFGS for training. If I randomize the initial weight, then it's normal if i always get different weight, right? Thanks for your response – MCRmy )

kjetil b halvorsen
  • 77,844