Skip to main content
2 of 2
added 76 characters in body

This approach may work well for points that are close to the training distribution. The prediction of the neural network for points outside the training distribution cannot be trusted a lot. It may happen that for points outside the training distribution the predictions of the neural network are somewhat strange and erratic. For more information on this topic have a look at this paper:

https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=1176122

Under the "IMPROPER BEHAVIOR OF MLP OUTSIDE THE BOUNDARY OF THE TRAINING SAMPLE" paragraph, you will find more information on this problem.