How can the number of hidden layers and numbers of hidden nodes be decided that will give optimal performance in a neural network?
(a) Using trial and error?
(b) using optimization Algorithms
(c) Is there any standard rule of thumb?
For Thumb rules
Heaton, J. (2008). The Number of Hidden Layers. Introduction to Neural Networks for Java. https://doi.org/10.1016/0893-6080(91)90009-T
For optimization, You have to tune parameters using your dataset.
Related to your query, I suggest you follow https://machinelearningmastery.com/how-to-configure-the-number-of-layers-and-nodes-in-a-neural-network/ and https://towardsdatascience.com/first-neural-network-for-beginners-explained-with-code-4cfd37e06eaf
For Thumb rules
Heaton, J. (2008). The Number of Hidden Layers. Introduction to Neural Networks for Java. https://doi.org/10.1016/0893-6080(91)90009-T
For optimization, You have to tune parameters using your dataset.
- Badges
- Science topic