Neural Network Architecture Beyond Width and Depth
This paper proposes a new neural network architecture by introducing an additional dimension called height beyond width and depth. Neural network architectures with height, width, and depth as hyperparameters are called three-dimensional architectures. It is shown that neural networks with three-dimensional architectures are significantly more expressive than the ones with two-dimensional architectures (those with only width and depth as hyperparameters), e.g., standard fully connected networks. The new network architecture is constructed recursively via a nested structure, and hence we call a network with the new architecture nested network (NestNet). A NestNet of height s is built with each hidden neuron activated by a NestNet of height ≤ s-1. When s=1, a NestNet degenerates to a standard network with a two-dimensional architecture. It is proved by construction that height-s ReLU NestNets with 𝒪(n) parameters can approximate Lipschitz continuous functions on [0,1]^d with an error 𝒪(n^-(s+1)/d), while the optimal approximation error of standard ReLU networks with 𝒪(n) parameters is 𝒪(n^-2/d). Furthermore, such a result is extended to generic continuous functions on [0,1]^d with the approximation error characterized by the modulus of continuity. Finally, a numerical example is provided to explore the advantages of the super approximation power of ReLU NestNets.
READ FULL TEXT