Papers
Topics
Authors
Recent
2000 character limit reached

Targeted materials discovery using Bayesian algorithm execution (2312.16078v1)

Published 26 Dec 2023 in cond-mat.mtrl-sci and physics.data-an

Abstract: Rapid discovery and synthesis of new materials requires intelligent data acquisition strategies to navigate large design spaces. A popular strategy is Bayesian optimization, which aims to find candidates that maximize material properties; however, materials design often requires finding specific subsets of the design space which meet more complex or specialized goals. We present a framework that captures experimental goals through straightforward user-defined filtering algorithms. These algorithms are automatically translated into one of three intelligent, parameter-free, sequential data acquisition strategies (SwitchBAX, InfoBAX, and MeanBAX). Our framework is tailored for typical discrete search spaces involving multiple measured physical properties and short time-horizon decision making. We evaluate this approach on datasets for TiO$_2$ nanoparticle synthesis and magnetic materials characterization, and show that our methods are significantly more efficient than state-of-the-art approaches.

Definition Search Book Streamline Icon: https://streamlinehq.com
References (62)
  1. J. B. Goodenough and K.-S. Park, “The li-ion rechargeable battery: a perspective,” Journal of the American Chemical Society 135, 1167–1176 (2013).
  2. P. A. Lee, N. Nagaosa, and X.-G. Wen, “Doping a mott insulator: Physics of high-temperature superconductivity,” Reviews of modern physics 78, 17 (2006).
  3. C. Suh, C. Fare, J. A. Warren, and E. O. Pyzer-Knapp, “Evolving the materials genome: How machine learning is fueling the next generation of materials discovery,” Annual Review of Materials Research 50, 1–25 (2020).
  4. J. H. Montoya, M. Aykol, A. Anapolsky, C. B. Gopal, P. K. Herring, J. S. Hummelshøj, L. Hung, H.-K. Kwon, D. Schweigert, S. Sun, et al., “Toward autonomous materials research: Recent progress and future challenges,” Applied Physics Reviews 9 (2022).
  5. B. Shahriari, K. Swersky, Z. Wang, R. P. Adams, and N. De Freitas, “Taking the human out of the loop: A review of bayesian optimization,” Proceedings of the IEEE 104, 148–175 (2015).
  6. A. G. Kusne, H. Yu, C. Wu, H. Zhang, J. Hattrick-Simpers, B. DeCost, S. Sarker, C. Oses, C. Toher, S. Curtarolo, et al., “On-the-fly closed-loop materials discovery via bayesian active learning,” Nature communications 11, 5966 (2020).
  7. M. J. Kochenderfer and T. A. Wheeler, Algorithms for optimization (Mit Press, 2019).
  8. A. Dave, J. Mitchell, K. Kandasamy, H. Wang, S. Burke, B. Paria, B. Póczos, J. Whitacre, and V. Viswanathan, “Autonomous discovery of battery electrolytes with robotic experimentation and machine learning,” Cell Reports Physical Science 1 (2020).
  9. M. T. Emmerich, A. H. Deutz, and J. W. Klinkenberg, “Hypervolume-based expected improvement: Monotonicity properties and exact computation,” in 2011 IEEE Congress of Evolutionary Computation (CEC) (IEEE, 2011) pp. 2147–2154.
  10. S. Daulton, M. Balandat, and E. Bakshy, “Differentiable expected hypervolume improvement for parallel multi-objective bayesian optimization,” Advances in Neural Information Processing Systems 33, 9851–9864 (2020).
  11. S. Daulton, M. Balandat, and E. Bakshy, “Parallel bayesian optimization of multiple noisy objectives with expected hypervolume improvement,” Advances in Neural Information Processing Systems 34, 2187–2200 (2021).
  12. J. Knowles, “Parego: A hybrid algorithm with on-line landscape approximation for expensive multiobjective optimization problems,” IEEE transactions on evolutionary computation 10, 50–66 (2006).
  13. A. Wang, H. Liang, A. McDannald, I. Takeuchi, and A. G. Kusne, “Benchmarking active learning strategies for materials optimization and discovery,” Oxford Open Materials Science 2, itac006 (2022).
  14. F. Hase, L. M. Roch, C. Kreisbeck, and A. Aspuru-Guzik, “Phoenics: a bayesian optimizer for chemistry,” ACS central science 4, 1134–1145 (2018).
  15. B. Rohr, H. S. Stein, D. Guevarra, Y. Wang, J. A. Haber, M. Aykol, S. K. Suram, and J. M. Gregoire, “Benchmarking the acceleration of materials discovery by sequential learning,” Chemical science 11, 2696–2706 (2020).
  16. T. Yamashita, N. Sato, H. Kino, T. Miyake, K. Tsuda, and T. Oguchi, “Crystal structure prediction accelerated by bayesian optimization,” Physical Review Materials 2, 013803 (2018).
  17. R. J. Hickman, M. Aldeghi, F. Häse, and A. Aspuru-Guzik, “Bayesian optimization with known experimental and design constraints for chemistry applications,” Digital Discovery 1, 732–744 (2022).
  18. H. C. Herbol, W. Hu, P. Frazier, P. Clancy, and M. Poloczek, “Efficient search of compositional space for hybrid organic–inorganic perovskites via bayesian optimization,” npj Computational Materials 4, 51 (2018).
  19. Y. Zhang, D. W. Apley, and W. Chen, “Bayesian optimization for materials design with mixed quantitative and qualitative variables,” Scientific reports 10, 4924 (2020).
  20. Q. Liang, A. E. Gongora, Z. Ren, A. Tiihonen, Z. Liu, S. Sun, J. R. Deneault, D. Bash, F. Mekki-Berrada, S. A. Khan, et al., “Benchmarking the performance of bayesian optimization across multiple experimental materials science domains,” npj Computational Materials 7, 188 (2021).
  21. B. J. Shields, J. Stevens, J. Li, M. Parasram, F. Damani, J. I. M. Alvarado, J. M. Janey, R. P. Adams, and A. G. Doyle, “Bayesian reaction optimization as a tool for chemical synthesis,” Nature 590, 89–96 (2021).
  22. F. Häse, M. Aldeghi, R. J. Hickman, L. M. Roch, and A. Aspuru-Guzik, “Gryffin: An algorithm for bayesian optimization of categorical variables informed by expert knowledge,” Applied Physics Reviews 8 (2021).
  23. K. Hanaoka, “Bayesian optimization for goal-oriented multi-objective inverse material design,” Iscience 24 (2021).
  24. M. Karasuyama, H. Kasugai, T. Tamura, and K. Shitara, “Computational design of stable and highly ion-conductive materials using multi-objective bayesian optimization: Case studies on diffusion of oxygen and lithium,” Computational Materials Science 184, 109927 (2020).
  25. B. Hu, Z. Wang, C. Du, W. Zou, W. Wu, J. Tang, J. Ai, H. Zhou, R. Chen, and B. Shan, “Multi-objective bayesian optimization accelerated design of tpms structures,” International Journal of Mechanical Sciences 244, 108085 (2023).
  26. W. Xu, Z. Liu, R. T. Piper, and J. W. Hsu, “Bayesian optimization of photonic curing process for flexible perovskite photovoltaic devices,” Solar Energy Materials and Solar Cells 249, 112055 (2023).
  27. X. Wang, Y. Huang, X. Xie, Y. Liu, Z. Huo, M. Lin, H. Xin, and R. Tong, “Bayesian-optimization-assisted discovery of stereoselective aluminum complexes for ring-opening polymerization of racemic lactide,” Nature Communications 14, 3647 (2023).
  28. D. Packwood et al., Bayesian optimization for materials science (Springer, 2017).
  29. M. M. Noack, K. G. Yager, M. Fukuto, G. S. Doerk, R. Li, and J. A. Sethian, “A kriging-based approach to autonomous experimentation with applications to x-ray scattering,” Scientific reports 9, 11809 (2019).
  30. N. J. Szymanski, C. J. Bartel, Y. Zeng, M. Diallo, H. Kim, and G. Ceder, “Adaptively driven x-ray diffraction guided by machine learning for autonomous phase identification,” npj Computational Materials 9, 31 (2023a).
  31. S. Ament, M. Amsler, D. R. Sutherland, M.-C. Chang, D. Guevarra, A. B. Connolly, J. M. Gregoire, M. O. Thompson, C. P. Gomes, and R. B. van Dover, “Autonomous materials synthesis via hierarchical active learning of nonequilibrium phase diagrams,” Science Advances 7, eabg4930 (2021).
  32. I. Bogunovic, J. Scarlett, A. Krause, and V. Cevher, “Truncated variance reduction: A unified approach to bayesian optimization and level-set estimation,” Advances in neural information processing systems 29 (2016).
  33. H. Ha, S. Gupta, S. Rana, and S. Venkatesh, “High dimensional level set estimation with bayesian neural network,” in Proceedings of the AAAI Conference on Artificial Intelligence, Vol. 35 (2021) pp. 12095–12103.
  34. K. Terayama, R. Tamura, Y. Nose, H. Hiramatsu, H. Hosono, Y. Okuno, and K. Tsuda, “Efficient construction method for phase diagrams using uncertainty sampling,” Physical Review Materials 3, 033802 (2019).
  35. C. Dai and S. C. Glotzer, “Efficient phase diagram sampling by active learning,” The Journal of Physical Chemistry B 124, 1275–1284 (2020).
  36. A. Y. Fong, L. Pellouchoud, M. Davidson, R. C. Walroth, C. Church, E. Tcareva, L. Wu, K. Peterson, B. Meredig, and C. J. Tassone, “Utilization of machine learning to accelerate colloidal synthesis and discovery,” The Journal of Chemical Physics 154, 224201 (2021).
  37. E. Y. Feng, R. Zelaya, A. Holm, A.-C. Yang, and M. Cargnello, “Investigation of the optical properties of uniform platinum, palladium, and nickel nanocrystals enables direct measurements of their concentrations in solution,” Colloids and Surfaces A: Physicochemical and Engineering Aspects 601, 125007 (2020).
  38. J. Rivnay, S. Inal, B. A. Collins, M. Sessolo, E. Stavrinidou, X. Strakosas, C. Tassone, D. M. Delongchamp, and G. G. Malliaras, “Structural control of mixed ionic and electronic transport in conducting polymers,” Nature communications 7, 11287 (2016).
  39. M. C. Prentiss, D. J. Wales, and P. G. Wolynes, “The energy landscape, folding pathways and the kinetics of a knotted protein,” PLoS computational biology 6, e1000835 (2010).
  40. A. R. Singh, J. H. Montoya, B. A. Rohr, C. Tsai, A. Vojvodic, and J. K. Nørskov, “Computational design of active site structures with improved transition-state scaling for ammonia synthesis,” ACS Catalysis 8, 4017–4024 (2018).
  41. N. Foloppe, L. M. Fisher, R. Howes, A. Potter, A. G. Robertson, and A. E. Surgenor, “Identification of chemically diverse chk1 inhibitors by receptor-based virtual screening,” Bioorganic & medicinal chemistry 14, 4792–4802 (2006).
  42. M. R. Palacín and A. de Guibert, “Why do batteries fail?” Science 351, 1253292 (2016).
  43. S. L. Scott, “A matter of life (time) and death,”  (2018).
  44. M. Jørgensen, K. Norrman, and F. C. Krebs, “Stability/degradation of polymer solar cells,” Solar energy materials and solar cells 92, 686–714 (2008).
  45. L. Di and E. H. Kerns, Drug-like properties: concepts, structure design and methods from ADME to toxicity optimization (Academic press, 2015).
  46. W. Neiswanger, K. A. Wang, and S. Ermon, “Bayesian algorithm execution: Estimating computable properties of black-box functions using mutual information,” in International Conference on Machine Learning (PMLR, 2021) pp. 8005–8015.
  47. S. A. Miskovich, W. Neiswanger, W. Colocho, C. Emma, J. Garrahan, T. Maxwell, C. Mayes, S. Ermon, A. Edelen, and D. Ratner, “Bayesian algorithm execution for tuning particle accelerator emittance with partial measurements,” arXiv preprint arXiv:2209.04587  (2022).
  48. R. Katsube, K. Terayama, R. Tamura, and Y. Nose, “Experimental establishment of phase diagrams guided by uncertainty sampling: an application to the deposition of zn–sn–p films by molecular beam epitaxy,” ACS Materials Letters 2, 571–575 (2020).
  49. J. A. G. Torres, P. C. Jennings, M. H. Hansen, J. R. Boes, and T. Bligaard, “Low-scaling algorithm for nudged elastic band calculations using a surrogate machine learning model,” Physical review letters 122, 156001 (2019).
  50. Y. Tian, R. Yuan, D. Xue, Y. Zhou, Y. Wang, X. Ding, J. Sun, and T. Lookman, “Determining multi-component phase diagrams with desired characteristics using active learning,” Advanced Science 8, 2003165 (2021).
  51. C. Tassone and A. Mehta, “Aggregation and structuring of materials and chemicals data from diverse sources,” Tech. Rep. (SLAC National Accelerator Lab., Menlo Park, CA (United States), 2019).
  52. Y. K. Yoo, Q. Xue, Y. S. Chu, S. Xu, U. Hangen, H.-C. Lee, W. Stein, and X.-D. Xiang, “Identification of amorphous phases in the fe–ni–co ternary alloy system using continuous phase diagram material chips,” Intermetallics 14, 241–247 (2006).
  53. V. Antonov, P. Oppeneer, A. Yaresko, A. Y. Perlov, and T. Kraft, “Computationally based explanation of the peculiar magneto-optical properties of ptmnsb and related ternary compounds,” Physical Review B 56, 13012 (1997).
  54. M. Abolhasani and E. Kumacheva, “The rise of self-driving labs in chemical and materials sciences,” Nature Synthesis , 1–10 (2023).
  55. B. P. MacLeod, F. G. Parlane, T. D. Morrissey, F. Häse, L. M. Roch, K. E. Dettelbach, R. Moreira, L. P. Yunker, M. B. Rooney, J. R. Deeth, et al., “Self-driving laboratory for accelerated discovery of thin-film materials,” Science Advances 6, eaaz8867 (2020).
  56. B. P. MacLeod, F. G. Parlane, C. C. Rupnow, K. E. Dettelbach, M. S. Elliott, T. D. Morrissey, T. H. Haley, O. Proskurin, M. B. Rooney, N. Taherimakhsousi, et al., “A self-driving laboratory advances the pareto front for material properties,” Nature communications 13, 995 (2022).
  57. N. J. Szymanski, B. Rendy, Y. Fei, R. E. Kumar, T. He, D. Milsted, M. J. McDermott, M. Gallant, E. D. Cubuk, A. Merchant, H. Kim, A. Jain, C. J. Bartel, K. Persson, Y. Zeng, and G. Ceder, “An autonomous laboratory for the accelerated synthesis of novel materials,” Nature  (2023b), 10.1038/s41586-023-06734-w.
  58. S. Chitturi, A. Ramdas, and W. Neiswanger, “src47/multibax-sklearn,”  (2023a).
  59. S. Chitturi, A. Ramdas, and W. Neiswanger, “src47/materials-bax-gpflow: Paper submission,”  (2023b).
  60. A. G. d. G. Matthews, M. van der Wilk, T. Nickson, K. Fujii, A. Boukouvalas, P. León-Villagrá, Z. Ghahramani, and J. Hensman, “GPflow: A Gaussian process library using TensorFlow,” Journal of Machine Learning Research 18, 1–6 (2017).
  61. V. Picheny, J. Berkeley, H. B. Moss, H. Stojic, U. Granta, S. W. Ober, A. Artemev, K. Ghani, A. Goodall, A. Paleyes, S. Vakili, S. Pascual-Diaz, S. Markou, J. Qing, N. R. B. S. Loka, and I. Couckuyt, “Trieste: Efficiently exploring the depths of black-box functions with tensorflow,”  (2023).
  62. M. H. et al, “python-ternary: Ternary plots in python,” Zenodo 10.5281/zenodo.594435 10.5281/zenodo.594435.

Summary

We haven't generated a summary for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Summarize Now

Whiteboard

Paper to Video (Beta)

Open Problems

We haven't generated a list of open problems mentioned in this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Generate Now

Continue Learning

We haven't generated follow-up questions for this paper yet.

Ai Sparkles Streamline Icon: https://streamlinehq.com Generate Now

Collections

Sign up for free to add this paper to one or more collections.

User Circle Single Streamline Icon: https://streamlinehq.com Sign Up