Metrics
82 Record Views
Conference proceeding
A Comparison of Machine Learning Techniques for the Generation of River and Stream Water Quality Estimates
ICAI 2014: PROCEEDINGS OF THE 2014 INTERNATIONAL CONFERENCE ON ARTIFICIAL INTELLIGENCE, Vol.2, pp.244-253
Volume 2
International Conference on Artificial Intelligence (ICAI), 16th (Las Vegas, NV, USA, 07/21/2014–07/24/2014)
01/01/2014
Abstract
Machine learning techniques can be used to classify water quality stations with similar concentration and discharge trends. Boosted Regression Trees (BRT) and Conditional Inference Trees (CIT) were considered as alternatives to conventional methods used by the United States Geological Survey to estimate daily concentrations of water constituents in rivers and streams based on continuous daily discharge data and discrete water quality samples collected at the same or nearby locations. The Weighted Regressions in Time, Discharge and Season (WRTDS) method is based on parametric survival regressions that generate unbiased estimates of the prediction errors. However, WRTDS needs a large number of samples collected during at least two decades. Alternatively, BRT and CIT can be used for water station classification by clustering data from nearby stations with similar concentration and discharge characteristics. This paper describes a machine learning tool that compares BRT, CIT, WRTDS, and clustering analysis for estimation of daily concentrations.
Details
- Title
- A Comparison of Machine Learning Techniques for the Generation of River and Stream Water Quality Estimates
- Publication Details
- ICAI 2014: PROCEEDINGS OF THE 2014 INTERNATIONAL CONFERENCE ON ARTIFICIAL INTELLIGENCE, Vol.2, pp.244-253
- Resource Type
- Conference proceeding
- Conference
- International Conference on Artificial Intelligence (ICAI), 16th (Las Vegas, NV, USA, 07/21/2014–07/24/2014)
- Publisher
- CSREA Press; United States
- Series
- Volume 2
- Format
- link
- Copyright
- Copyright © 2014 CSREA Press
- Identifiers
- 99380461297306600
- Academic Unit
- Office of Teaching, Learning, and Technology; College of Arts, Social Sciences, and Humanities; Center for Cybersecurity and AI
- Language
- English