Premium
Using Decision Trees to Predict Drinking Water Advisories in Small Water Systems
Author(s) -
Murphy Heather M.,
Bhatti Munir,
Harvey Richard,
McBean Edward A.
Publication year - 2016
Publication title -
journal ‐ american water works association
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.466
H-Index - 74
eISSN - 1551-8833
pISSN - 0003-150X
DOI - 10.5942/jawwa.2016.108.0008
Subject(s) - water source , decision tree , water use , water infrastructure , environmental planning , geographic information system , environmental resource management , environmental science , computer science , water resource management , geography , water supply , environmental engineering , data mining , ecology , cartography , biology
As of Jan. 1, 2015, there were 1,838 drinking water advisories (DWAs) in effect across Canada, including DWAs in First Nations communities. This research investigates the use of data‐mining techniques to identify which factors can potentially lead to a DWA in small water systems such as those found in First Nations communities in Canada. The results show that the training level of operators, remoteness/geographic location, source water type, and the class of treatment system are factors that influence whether a DWA is issued in a water system. The decision trees discussed in this study demonstrate that data mining is capable of correctly predicting up to 79% of future DWAs. This study demonstrates that a decisiontree methodology is a powerful, user‐friendly tool that can help water managers and regulators better understand vulnerabilities related to the provision of drinking water in small systems.