1. Title, Journal and Authors
- Title: Predicting cyanobacterial abundance, microcystin, and geosmin in a eutrophic drinking-water reservoir using a 14-year dataset
- Journal : Lake and Reservoir Management
- Authors : Ted D. Harris, Jennifer L. Graham
- There have been numerous attempts to create models that predict cyanobacteria and their secondary metabolites, most using linear models.
- However, linear models are limited by assumptions about the data
- This articles makes non-linear models using Support Vector Machine(SVM), Random Forest(RF), Boosted Tree(BT), Cubist modeling techniques.
- Cross validation is used for comparison between models.
- In the wide range of cyanobacteria concentration, Random Forest(RT) shows the highest performance for prediction.
- Considering the only high range of cyanobacteria, cubist model shows the best performance
3. Application to research
- Non-linear techniques, especially RT and cubist models can be used to predict the cyanobacteria
- If the purpose of the prediction is detecting the 'alert range of cyanobacteria concentration', then the cubist model can be the alternative choice for prediction.
Sungho Shin (Ph.D. program)
Environmental Systems Engineering Lab.
School of Environmental Science & Engineering
Gwangju Institute of Science and Technology
1 Oryong-dong Buk-gu Gwangju, 500-712, Korea
Email : firstname.lastname@example.org