This project provides research opportunities  to develop statistical methodology or computer software to address increasing needs at NIST on using machine learning to solve interesting engineering applications or physical measurement problems. An example is the cooktop fire experiments where multiple sensors are used to monitor emissions from cooking and the goal is to select sensors and sensor signatures to alert eminent fire. The time series research issues include nonparametric derivative estimation and classification techniques applied to multivarite time-varying signals. A similar issue arises in a medical imaging problem where tissue images are collected using high resolution and hyperspectral image sensors and the goal is to use deep learning or other statistical learning technqiues such as matrix singular value decomposition to extract spectral signatures that can distinguish abnormal from normal tissues and help surgeon doctors to mark up the tissue boundary more accurately and quickly.  

References: 

A.E. Mensch, A.P. Hamins, J. Lu, W.C. Tam (Fire Technology 2021), 

Sensors and Machine Learning Models to Prevent Cooktop Ignition and Ignore Normal Cooking.