I have a huge dataset. Input length is 5038 float numbers and output is 6690 binary numbers. Can someone suggest the number of hidden layers and nodes? I am working in TensorFlow.
There is no hard and fast rule to decide for the number of hidden layers and no. of nodes. The optimal number of hidden units could easily be smaller than the number of inputs, there is no rule like multiply the number of inputs with N. You can use cross validation to check the accuracy on test set and accordingly you can change your hidden units to get good accuracy.