Dr. Calderon has been involved with developing new statistics and machine learning applications for uses in the basic physical and life sciences as well as industrial applications (industrial efforts at his company, Ursa Analytics, are focused on signal processing and target tracking). The common theme underlying the various applications are data-driven modeling and hypothesis testing associated with noisy, correlated data (images and time series data). He is particularly interested in interdisciplinary research collaboration aiming to leverage advances in machine learning and statistics to analyze images and signals recorded in nano to micro scale experimental systems. Applications have included time series analysis of single-molecule measurements to characterize biomolecules work in both in vivo and in vitro environments, supervised and unsupervised machine learning/image analysis applied to high-throughput devices, and analysis of noisy trace chemical detector signals in agricultural and military applications.
