LGATMLJul 26, 2019

Two-hidden-layer Feedforward Neural Networks are Universal Approximators: A Constructive Approach

arXiv:1907.11457v2
Originality Synthesis-oriented
AI Analysis

This offers a constructive method for universal approximation, addressing a theoretical gap in neural network theory, though it is incremental as it builds on existing existence proofs.

The paper tackles the problem of constructing a two-hidden-layer feedforward neural network to approximate continuous functions, providing a concrete architecture and weights based on simplicial maps, with approximation quality depending on the refinement of the space covering.

It is well known that Artificial Neural Networks are universal approximators. The classical result proves that, given a continuous function on a compact set on an n-dimensional space, then there exists a one-hidden-layer feedforward network which approximates the function. Such result proves the existence, but it does not provide a method for finding it. In this paper, a constructive approach to the proof of this property is given for the case of two-hidden-layer feedforward networks. This approach is based on an approximation of continuous functions by simplicial maps. Once a triangulation of the space is given, a concrete architecture and set of weights can be obtained. The quality of the approximation depends on the refinement of the covering of the space by simplicial complexes.

Foundations

The foundational work for this paper's niche, ranked by how specifically the neighbourhood builds on it — not by global fame.

Your Notes