Deep neural network has been proved that efficient in imaging, text and voice data. Are they work good in processing low dimensional data ? (One example, using the DNN to analysis mechanical reliability problem whose limit state function (LSF) has only several variables (low dimension)).
This discussion might be useful;
https://www.quora.com/Are-deep-neural-networks-useful-for-problems-with-low-dimensional-input
Dear Houzhu Ding
It could be possible, depending of the nature of your problem, but you could use other neural network techniques like radial based neural networks. The training of these nets, treaty as an inverse problem, allows diminishing the generalized errors when applying regularization technics.
I hope my answer will be useful for you.
Best wishes,
Jose Arzola
for big data problems, deep NN can be most appropriate, However for low dimensional data other NNs can be used.
- Badges
- Science topic
- Similar topics
- Engineering
- Computer Engineering