Adaptive active subspace-based efficient multifidelity materials design

Bayesian optimization with active learning of design constraints using an entropy-based approach

“…σ y ðT; _ ϵÞ ¼ Mτ y 0 (13) The variables in the above equation are as follows: α is the line tension parameter and is set to 1/12 for edge dislocations; μ is the average shear modulus of the alloy; ν is the average Poisson ratio of the alloy; b is the Burger vector associate with the BCC edge dislocation within the random alloy; ΔV is the misfit volume of the nth solute, which can be accurately estimated as ΔV n = V n − ∑ n =1c n V n according to Vegard's law; τ y 0 is the zero-temperature yield stress; ΔE b is the energy barrier for the thermal-activated flow; _ ϵ 0 is the reference strain rate which is typically set to 10 4 s −1 ; _ ϵ is the applied strain rate which is typically set to 10 3 s −1 and is indeed set to this value in the current work; M is the Taylor factor for edge glide in a random BCC polycrystal; k B is the Boltzmann constant; σ y (T, ϵ) is the yield strength estimated at a finite temperature and strain rate, T and _ ϵ. In this study, we employed the DFT-based KKR (Korringa-Kohn-Rostoker Green's function) method as the reference model for calculating key properties such as intrinsic strength and phase stability for arbitrary compositions.…”

“…An improvement to the Bayesian optimization paradigm is to employ multiple models representing the same quantity of interest. This is known as multi-fidelity BO and has been shown to effectively increase the robustness and efficiency of engineering design schemes [12][13][14][15][16] . These models are built upon different assumptions and/or simplifications and vary in fidelity and cost of the evaluation.…”

“…In the earlier works of refs. [12][13][14][15][16] , a multi-fidelity approach has been employed to optimize a single quantity of interest (single-objective optimization). Recently, this multi-fidelity setting has been expanded to multi-objective design problems as well 17 .…”

Computational and data-driven modelling of solid polymer electrolytes

“…102 Besides applications in so nanomaterials, BO has also been extensively applied in energy storage materials, 100 microstructures of nanomechanical resonators, 108 and alloy design using multi-delity approaches. 109,110 From the optimization process perspective, Nakayama et al surveyed the use of acquisition functions and initial values in the BO materials synthesis as a simplied 1D case. 111 Bellamy et al used batch BO to explore a large database for use in drug design.…”

Machine learning-assisted design of material properties

“…Feature importance analysis can be performed intermittently to eliminate design parameters that only marginally influence the property, thus reducing the number of dimensions to explore in the subsequent iterations. 100,101 Different techniques for active learning can be categorized based on the choice of ML model used to predict the property landscape and the iterative algorithm employed for determining the next design points to probe. Bayesian optimization is an active learning algorithm widely discussed in previous review articles 30-32, 110, 111 which fits a Gaussian process regression model to the labeled data points at every iteration.…”