Classification of Undetermined Deaths by Poisoning: Comparison of Homogeneous and Heterogeneous Databases

Authors

  • Alana Esty Systems and Computer Engineering, Carleton University
  • Monique Frize Systems and Computer Engineering, Carleton University School of Electrical Engineering and Computer Science, University of Ottawa
  • Steven R. McFaull Surveillance and Epidemiology Division, Public Health Agency of Canada, Ottawa
  • Robin Skinner Surveillance and Epidemiology Division, Public Health Agency of Canada, Ottawa
  • Jenini Subaskaran Surveillance and Epidemiology Division, Public Health Agency of Canada, Ottawa
  • Melinda Tiv Surveillance and Epidemiology Division, Public Health Agency of Canada, Ottawa

Abstract

Identifying factors that differentiate suicides from unintentional deaths by poisoning is essential for accurate monitoring of suicide in order to help develop prevention measures. The results of our research demonstrated that the use of machine learning techniques such as: Artificial Neural Networks, Decision Trees, and Case-Based Reasoning have enabled us to classify the majority of undetermined cases found in the two databases analyzed. The data originated from the Canadian Coroner and Medical Examiner Database (CCMED): the first dataset included deaths in the Province of Ontario; the second dataset included cases from several other provinces excluding Ontario.

Downloads

Published

2017-05-23

How to Cite

[1]
A. Esty, M. Frize, S. R. McFaull, R. Skinner, J. Subaskaran, and M. Tiv, “Classification of Undetermined Deaths by Poisoning: Comparison of Homogeneous and Heterogeneous Databases”, CMBES Proc., vol. 40, no. 1, May 2017.

Issue

Vol. 40 (2017)

Section

Academic