













Abstract:We show that deep binary ReLU networks and deep sparse ternary ReLU networks can approximate $\beta$-Hölder functions on $[0,1]^d$. We also show that deep sparse ternary ReLU networks can achieve, up to a logarithmic factor, the minimax rate of prediction of $\beta$-smooth regression function.
From: Aleksandr Beknazaryan [view email]
[v1]
Mon, 27 Jun 2022 13:16:08 UTC (8 KB)
[v2]
Tue, 28 Jun 2022 04:45:45 UTC (8 KB)
[v3]
Fri, 8 Jul 2022 11:37:35 UTC (8 KB)
[v4]
Mon, 10 Aug 2026 12:05:13 UTC (8 KB)
此内容由惯性聚合(RSS阅读器)自动聚合整理,仅供阅读参考。 原文来自 — 版权归原作者所有。