notesum.ai

Published at November 2

Supervised Score-Based Modeling by Gradient Boosting

cs.LG

cs.AI

Released Date: November 2, 2024

Authors: Changyuan Zhao¹, Hongyang Du², Guangyuan Liu¹, Dusit Niyato¹

Aff.: ¹College of Computing and Data Science, Nanyang Technological University; ²Department of Electrical and Electronic Engineering, University of Hong Kong

Arxiv: http://arxiv.org/abs/2411.01159v1

	RMSE $\downarrow$
	$\lambda(\sigma_{i})=\sigma_{i}$ (L1)				$\lambda(\sigma_{i})=\sigma_{i}^{2}$ (L2)
	w/o noise, fast	w/ noise, fast	w/ noise	w/o noise	w/ noise, fast	w/o noise	w/o noise, fast	w/ noise
Linear	0.0667	$0.0668$	$0.1767$	$0.0773$	$0.0685$	$0.0822$	$0.0684$	$0.1764$
Quadratic	$0.1197$	$0.1198$	$0.1103$	$0.1296$	$0.0951$	$0.1026$	0.0950	$0.1139$
Log-Log Linear	$0.0783$	$0.0781$	$0.0857$	0.0654	$0.0793$	$0.0885$	$0.0794$	$0.0699$
Log-Log Cubic	$3.2704$	$3.2213$	$3.3903$	$3.9410$	1.9919	$2.9880$	$2.0050$	$3.2718$
Sinusoidal	$0.0096$	$0.0097$	$0.0098$	$0.0094$	$0.0035$	$0.0033$	0.0033	$0.0034$