Composite Bayesian Optimization In Function Spaces Using NEON -- Neural Epistemic Operator Networks

arXiv:2404.03099v1 Announce Type: new
Abstract: Operator learning is a rising field of scientific computing where inputs or outputs of a machine learning model are functions defined in infinite-dimensional spaces. In this paper, we introduce NEON (Neural Epistemic Operator Networks), an architecture for generating predictions with uncertainty using a single operator network backbone, which presents orders of magnitude less trainable parameters than deep ensembles of comparable performance. We showcase the utility of this method for sequential decision-making by examining the problem …

abstract architecture arxiv bayesian computing cs.ai cs.ce cs.it cs.lg function functions inputs machine machine learning machine learning model math.it neon networks optimization paper predictions scientific spaces stat.ml type uncertainty