Application of Deep Learning Neural Networks for Nitrate Prediction in the Klokot River, Bosnia and Herzegovina

Abstract

Nitrate is one of the focal water quality indices in aquatic systems. However, proper estimation of nitrate concentration is a complicated task. In this article, capabilities of deep neural networks (DNN) and conventional artificial neural networks (ANN) to model and predict nitrate concentration in the Klokot River, Bosnia and Herzegovina were investigated. The measured data includes nitrate and pH values. Different scenarios were considered as the potential models for DNN and ANN structuring. The result showed DNN and ANN networks are incapable of precisely modeling the nitrate concentration in the Klokot River based on the designed scenarios. In addition, DNN was slightly superior to ANN in different terms of estimation accuracy.