Abstract:Deeper neural networks are more difficult to train. We present a residual
learning framework to ease the training of networks that are substantially
deeper than those used previously. We explicitly reformulate the layers as
learning residual functions with reference to the layer inputs, instead of
learning unreferenced functions. We provide comprehensive empirical evidence
showing that these residual networks are easier to optimize, and can gain
accuracy from considerably increased depth. On the ImageNe ...
The authors have found that residual networks can increase accuracy by significantly increasing depth. In one of their experiments they found that a 1202-layer network performed worse than a 110-layer network showing there is some limit to the gain in accuracy that is achieved by increasing depth.
Given the number of nodes and the size of the dataset is there an approximate function which tells you how many layers you should have to optimize the accuracy of the neural network?