Autonomous Materials Discovery Driven by Gaussian Process Regression\n with Inhomogeneous Measurement Noise and Anisotropic Kernels

2020/06/03 by Marcus M. Noack, Gregory S. Doerk, Noack, Marcus M. +11 · 3 citations
Biochemistry, Genetics and Molecular Biology · Materials Science · #Computational Physics (physics.comp-ph) #FOS: Computer and information sciences #FOS: Physical sciences #Machine Learning (stat.ML) #Machine Learning in Materials Science #Molecular Biology Techniques and Applications #Spectroscopy Techniques in Biomedical and Chemical Research

paper · pdf · doi:10.48550/arxiv.2006.02489

openalex publication_date 2020/06/03 · openalex created_date 2022/07/26 · openalex updated_date 2026/07/28

Abstract

A majority of experimental disciplines face the challenge of exploring large\nand high-dimensional parameter spaces in search of new scientific discoveries.\nMaterials science is no exception; the wide variety of synthesis, processing,\nand environmental conditions that influence material properties gives rise to\nparticularly vast parameter spaces. Recent advances have led to an increase in\nefficiency of materials discovery by increasingly automating the exploration\nprocesses. Methods for autonomous experimentation have become more\nsophisticated recently, allowing for multi-dimensional parameter spaces to be\nexplored efficiently and with minimal human intervention, thereby liberating\nthe scientists to focus on interpretations and big-picture decisions. Gaussian\nprocess regression (GPR) techniques have emerged as the method of choice for\nsteering many classes of experiments. We have recently demonstrated the\npositive impact of GPR-driven decision-making algorithms on autonomously\nsteering experiments at a synchrotron beamline. However, due to the complexity\nof the experiments, GPR often cannot be used in its most basic form, but rather\nhas to be tuned to account for the special requirements of the experiments. Two\nrequirements seem to be of particular importance, namely inhomogeneous\nmeasurement noise (input dependent or non-i.i.d.) and anisotropic kernel\nfunctions, which are the two concepts that we tackle in this paper. Our\nsynthetic and experimental tests demonstrate the importance of both concepts\nfor experiments in materials science and the benefits that result from\nincluding them in the autonomous decision-making process.\n

Cited by

• Noise-Aware Optimization in Nominally Identical Manufacturing and Measuring Systems for High-Throughput Parallel Workflows
• Inverse design of drying-induced assembly of multicomponent colloidal-particle films using surrogate models
• Human-AI Synergy in Adaptive Active Learning for Continuous Lithium Carbonate Crystallization Optimization

Related

• Physics makes the difference: Bayesian optimization and active learning via augmented Gaussian process
• Active oversight and quality control in standard Bayesian optimization for autonomous experiments
• A Data-Driven Method for Automated Data Superposition with Applications in Soft Matter Science
• Gaussian Process Regression for Materials and Molecules
• Accelerated Materials Discovery through Cost-Aware Bayesian Optimization of Real-World Indentation Workflows