Authors: Brent Vlodarchyk*, , Chris Houser, University of Windsor , Jacob Lehner , University of Windsor , Phil Wernette, University of Windsor
Topics: Hazards and Vulnerability, Spatial Analysis & Modeling, Human-Environment Geography
Keywords: Machine Learning, Drowning, Great Lakes
Session Type: Poster
Start / End Time: 8:00 AM / 9:40 AM
Room: Napoleon Foyer/Common St. Corridor, Sheraton, 3rd Floor
Presentation File: No File Uploaded
Drownings on the Great Lakes are a public health issue in both Canada and the United States. Reducing the number of drownings is complicated by the fact that the number of drownings vary year to year with little consistency. This study examines the spatial and temporal variation of drownings on the Great lakes between 2010 and 2017 in order to identify common factors among drowning events, as well as surf hazards. Examining weather and climatic factors and the demographics of each drowning. Specifically, GIS (Geographic Information System) is used to show the spatial and temporal variation in the drownings and, to serve as a database to develop machine learning prediction program. A total of 391 drownings occurred on the Great Lakes between 2010 and 2017 and, further analyses suggest that temperature, wind speed, and wind direction are important predictors of drownings for particular user groups (based on age, gender and location). Temperature and the number of drownings are positively correlated, with the most drownings occurring during years that have the highest temperature. Ice concentration was also found to have a strong correlation to the number of drownings each year. This can be used as a prediction marker for the number of drownings likely to occur in the upcoming summer. These factors will be used as variables to be used in machine learning based analysis to predict the number of drownings to occur each year.