Marc Berghouse

Background

About Marc Berghouse

I am a fourth year hydrology PhD student at the University of Nevada, Reno (projected to finish in October, 2024). My PhD research lies at the intersection of microbial motility, fluid dynamics and machine learning. One primary aspect of my research is the use particle tracking to characterize the transport of bacteria at the micro scale as a function of variable speed, porosity, and motility type. I then plan to upscale these results with the reactive transport simulators STOMP and PFLOTRAN to investigate the field scale impact of microbial motility on bioremediation. I am also using PFLOTRAN to conduct research on hotspot formation in the hyporheic zone as a function of variable permeability, nutrient concentrations and flow conditions. In addition to the physics-based aspects of my PhD research, I am also pursuing the publication of multiple papers on using deep learning to enhance various aspects of particle tracking and reactive transport simulation workflows. Besides my PhD research, I have been been coding in Python and working on data science problems for the last 8 years. My first presentation of deep learning research was at the KGML conference in 2021, where I used a physics guided neural net to predict CO2 concentrations in the soil. Currently, in addition to the deep learning aspects of my PhD research, I'm working on a project that seeks to compare attention methods in CNN's for breast cancer classification in mammograms. I published this research at ISVC conference in October, 2023, and a follow up paper is currently under peer review in Image and Vision Computing. In addition, during the Spring 2023 semester I created and co-taught a graduate level class at UNR on the applications of machine learning in hydrology. https://marcberghouse.wixsite.com/website https://twitter.com/BerghouseMarc https://github.com/mberghouse https://hub.docker.com/u/mberghouse