0

I am training Autoencoder (1 hl) on keras. I have used validation split of 20%. Training seems ok, but for the validation set the MSE error(shown in blue) seems to saturate with whatever no. of hidden nodes. It does not goes further down. What could be possible reason? enter image description here

I've tried various LR with adam. At 0.001, there are kinks in graph, at 0.0001(this one) the graph is smooth but still saturates.

Bully Maguire
  • 121
  • 2
  • Does this answer your question? [What should I do when my neural network doesn't learn?](https://stats.stackexchange.com/questions/352036/what-should-i-do-when-my-neural-network-doesnt-learn) – Stephan Kolassa Feb 08 '22 at 16:07
  • doesn't seem like my model isnt learning. The training error is going only down – Bully Maguire Feb 08 '22 at 16:09
  • Sorry, you are right, misclick. How about [How to know that your machine learning problem is hopeless?](https://stats.stackexchange.com/q/222179/1352)? – Stephan Kolassa Feb 08 '22 at 16:13

0 Answers0