Wednesday, October 2nd , 2019

Speaker: Dr. Paul Hynds, Principal Investigator, Technological University Dublin.

Title: “Random Forests and Kidney Failure: A statistical journey toward predicting pediatric Hemolytic Uremic Syndrome in the Republic of Ireland”

Time: 2:30 – 3:30 PM

Location: Rm. 217, Dupuis Hall, Queen’s University

Refreshments provided!

 

Bio

Dr. Paul Hynds is a Principal Investigator in the Environmental Sustainability and Health Institute (ESHI), Technological University Dublin, having graduated with PhDs in Environmental Engineering in 2012 and Mathematical Statistics in 2015, both from Trinity College Dublin. His primary interests lie in environmental fate modelling, environmentally acquired infections and statistical epidemiology; most of his work is borne out of the "dual receptor" model thus acknowledging that both the environment and human are affected by contamination, and particularly pathogenic contamination from multiple sources. As such, he tries to predict the presence, frequency, and movement of enteric pathogens in the subsurface environment, and how, when, where and among whom these pathogens cause infectious diseases. Current projects include WELLness, GRAppLE, ISO-MECH, SMARTIE, STEP-WISE and EPI-CENTRE. He also loves climbing, rugby and acronyms. 

 

Abstract

The seminar will present multiple multi-disciplinary studies, all of which had and have a very similar overarching objective, but use several approaches, datasets, test sites, and software packages to get there. Ireland currently has the highest incidence rate of verotoxigenic E. coli enteritis in Europe and one of the highest in the world. For example, current incidence rates for Ireland and Canada are approximately 9 cases/100,000 persons and 1.8 cases/100,000 persons, respectively. Moreover, the profile of infection in Ireland is suspected to be extremely unique due to it’s climate, infrastructure, and geography. Thus, the “journey” comprises hydrogeology, microbial ecology, socio-epidemiology, statistical modelling, machine learning, quantitative risk assessment, and Paul getting stuck down a well for 4 hours (True Story).