TY  - JOUR
TI  - An Artificial Neural Network Model for Water Quality and Water Consumption Prediction
ID  - uneatlantico5425
KW  - water quality prediction; water consumption prediction; artificial neural network; classification
A1  - Rustam, Furqan
A1  - Ishaq, Abid
A1  - Kokab, Sayyida Tabinda
A1  - de la Torre Diez, Isabel
A1  - Vidal Mazón, Juan Luis
A1  - Rodríguez Velasco, Carmen Lilí
A1  - Ashraf, Imran
AV  - public
IS  - 21
Y1  - 2022/10//
JF  - Water
N2  - With rapid urbanization, high rates of industrialization, and inappropriate waste disposal, water quality has been substantially degraded during the past decade. So, water quality prediction, an essential element for a healthy society, has become a task of great significance to protecting the water environment. Existing approaches focus predominantly on either water quality or water consumption prediction, utilizing complex algorithms that reduce the accuracy of imbalanced datasets and increase computational complexity. This study proposes a simple architecture of neural networks which is more efficient and accurate and can work for predicting both water quality and water consumption. An artificial neural network (ANN) consisting of one hidden layer and a couple of dropout and activation layers is utilized in this regard. The approach is tested using two datasets for predicting water quality and water consumption. Results show a 0.96 accuracy for water quality prediction which is better than existing studies. A 0.99 R2 score is obtained for water consumption prediction which is superior to existing state-of-the-art approaches.
UR  - http://doi.org/10.3390/w14213359
SN  - 2073-4441
VL  - 14
ER  -