ARCHIVES
VOL. 6, ISSUE 3 (2024)
Development of a machine learning predictive model for identification of key water quality pollution parameters at River Benue
Authors
GA Mohamed, MO Udochukwu, Dr. SO Enokela, YD Kantiok
Abstract
This study on development of a machine learning predictive model for
identification and classification of water pollution parameters at River Benue
at Makurdi reach applied relevant artificial intelligence models to verify the
pollution levels of the river. The water samples were collected in sterile
bottles of 1,500 mL capacity at the depth of 20 cm and examined in the
laboratory using standard methods. Results of the verified laboratory data sets
tested on the different AI models were used to validate the modal as to
identify the model that best predict water pollution levels of river Benue in
terms of EC, SAR, SO4, TDS relevant to irrigation water quality
standards. The Ensemble (Bagged Tree), Ensemble (Boosted Tree) and the SVM
(Quadratic) models emerges as the optimal predictive model for estimating SAR,
DO and TDS respectively in River Benue at Makurdi. The observed and simulated
values in the form of scatter plots were used to validate the model. The
scatter plots reported an acceptable deviation from the ideal line of the 45o,
which confirm the degree of the correlation between the observed and simulated
dataset. An additional analysis was performed using the SI variations of
SVM-FFA model “as a superior predictive model” among different test stages.
Download
Pages:97-106
How to cite this article:
GA Mohamed, MO Udochukwu, Dr. SO Enokela, YD Kantiok "Development of a machine learning predictive model for identification of key water quality pollution parameters at River Benue". International Journal of Ecology and Environmental Sciences, Vol 6, Issue 3, 2024, Pages 97-106
Download Author Certificate
Please enter the email address corresponding to this article submission to download your certificate.