Approximation Error and Complexity Bounds for ReLU Networks on Low-Regular Function Spaces
This work provides theoretical bounds for neural network approximation, which is incremental as it builds on existing Fourier features residual networks.
The paper tackles the problem of approximating bounded functions with minimal regularity assumptions using ReLU neural networks, showing that the approximation error is bounded by a quantity proportional to the uniform norm of the target function and inversely proportional to the product of network width and depth.
In this work, we consider the approximation of a large class of bounded functions, with minimal regularity assumptions, by ReLU neural networks. We show that the approximation error can be bounded from above by a quantity proportional to the uniform norm of the target function and inversely proportional to the product of network width and depth. We inherit this approximation error bound from Fourier features residual networks, a type of neural network that uses complex exponential activation functions. Our proof is constructive and proceeds by conducting a careful complexity analysis associated with the approximation of a Fourier features residual network by a ReLU network.