Authors: Abolfazl Mollalo*, Baldwin Wallace University
Topics: Geographic Information Science and Systems, Environmental Science, Spatial Analysis & Modeling
Keywords: Machine learning; Medical Geography; Spatial Statistics; GIS
Session Type: Poster
Presentation File: No File Uploaded
In this study we integrated spatial statistical tools with machine learning classifiers in a GIS platform to predict presence/absence of the hotspots of age-adjusted lower respiratory infection (LRI) mortality rate across the continental US from 1980 to 2014. Explaining how the counties with high LRI mortality rates cluster geographically can be helpful to further reduce LRI mortality rates. Moreover, results of machine learning classifiers including random forest, support vector machine, k-Nearest neighbour and logistic regression in predicting presence/absence of hotspots can provide additional insight for future researches aimed at understanding what can be contributing to higher than expected LRI mortality rates. Findings of this study can be used by public health decision-makers to the development and implementation of LRI prevention efforts by facilitating better targeting of resources toward areas in greatest need.
To access contact information login