Emulation of physical processes with Emukit

Abstract

Decision making in uncertain scenarios is an ubiquitous challenge in real world systems. Tools to deal with this challenge include simulations to gather information and statistical emulation to quantify uncertainty. The machine learning community has developed a number of methods to facilitate decision making, but so far they are scattered in multiple different toolkits, and generally rely on a fixed backend. In this paper, we present Emukit, a highly adaptable Python toolkit for enriching decision making under uncertainty. Emukit allows users to: (i) use state of the art methods including Bayesian optimization, multi-fidelity emulation, experimental design, Bayesian quadrature and sensitivity analysis; (ii) easily prototype new decision making methods for new problems. Emukit is agnostic to the underlying modeling framework and enables users to use their own custom models. We show how Emukit can be used on three exemplary case studies.

arXiv commentSecond Workshop on Machine Learning and the Physical Sciences, NeurIPS 2019

References in corpus (4)

• Coupling Superconducting Qubits via a Cavity Bus
• Virtual vs. Real: Trading Off Simulations and Physical Experiments in Reinforcement Learning with Bayesian Optimization
• GPflowOpt: A Bayesian Optimization Library using TensorFlow
• Survey of multifidelity methods in uncertainty propagation, inference, and optimization

Cited by in corpus (18)

• Bayesian Optimization is Superior to Random Search for Machine Learning Hyperparameter Tuning: Analysis of the Black-Box Optimization Challenge 2020
• Efficient emulation of relativistic heavy ion collisions with transfer learning
• Cosmological Constraints from the eBOSS Lyman-$\alpha$ Forest using the PRIYA Simulations
• BOSS: Bayesian Optimization over String Spaces
• Multi-fidelity data fusion through parameter space reduction with applications to automotive engineering
• Graph Neural Network Guided Local Search for the Traveling Salesperson Problem
• PyVBMC: Efficient Bayesian inference in Python
• GIBBON: General-purpose Information-Based Bayesian OptimisatioN
• Optimised one-class classification performance
• Multi-fidelity data fusion for the approximation of scalar functions with low intrinsic dimensionality through active subspaces
• Transfer Learning with Gaussian Processes for Bayesian Optimization
• Overview of Gaussian process based multi-fidelity techniques with variable relationship between fidelities
• BOSH: Bayesian Optimization by Sampling Hierarchically
• Good practices for Bayesian Optimization of high dimensional structured spaces
• Constrained, Global Optimization of Functions with Lipschitz Continuous Gradients
• Computationally-Efficient Climate Predictions using Multi-Fidelity Surrogate Modelling
• Sonic: A Sampling-based Online Controller for Streaming Applications
• GaussED: A Probabilistic Programming Language for Sequential Experimental Design