Error bounds for deep ReLU networks using the Kolmogorov--Arnold superposition theorem
Columbia University（哥伦比亚大学）；National University of Singapore（新加坡国立大学）
英文摘要：
We prove a theorem concerning the approximation of multivariate functions by deep ReLU networks, for which the curse of the dimensionality is lessened. Our theorem is based on a constructive proof of the Kolmogorov--Arnold superposition theorem, and on a subset of multivariate continuous functions whose outer superposition functions can be efficiently approximated by deep ReLU networks.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
