Underlying the astonishing results and growing applications of artificial intelligence (AI) are fundamental concepts from statistics, mathematics, and data science. This post-doctoral research opportunity seeks a candidate well-versed and eager to grow in these areas; the candidate would gain not only valuable experience in machine learning (ML) but would collaborate and publish with NIST staff on the applications of ML to an industrially relevant, optics-based measurement challenge.
Specifically, we determine the sizes of objects seized well below conventional optical scattering limits using imaging and rigorous electromagnetic simulations. Determining such dimensions from optical scattering is an inverse problem that is often solved by comparing libraries of simulated curves against experimental measurements using nonlinear regression; this regression not only yields values for these parameters but also their uncertainties, which is essential to quantitative measurements.
Not only might AI enhance our capabilities to measure subwavelength features, we envision AI will be required to process optical measurements in the near future for nanoelectronics due to increasing architectural and materials complexity. The grand challenge faced is that there is currently a disconnect between the remarkable results from ML and an accurate understanding of the uncertainty of such results. Note that this challenge extends beyond optical measurements as it is critical to incorporating the potential of AI into measurement science.
References:
M.A. Henn, H. Zhou, B. M. Barnes, "Data-driven approaches to optical patterned defect detection," OSA Continuum, Vol. 2, No. 9, pp. 2683-2693 (2019).
N. G. Orji, M. Badaroglu, B. M. Barnes, C. Beitia, B. D. Bunday, U. Celano, R. J. Kline, M. Neisser, Y. S. Obeng, A. Vladar, "Metrology for the next generation of semiconductor devices," Nature Electronics, Vol. 1, No. 10, pp. 532-547 (2018).
Artificial intelligence; machine learning; applied mathematics; uncertainty; accuracy; optics; measurement; metrology; variance; defect metrology;