Multiple Statistical Model Ensemble Predictions of Residual Chlorine in Drinking Water: Applications of Various Deep Learning and Machine Learning Algorithms

Abstract

Applicability of statistical models in predicting chlorine decay remains minimally explored. This study predicted residual chlorine using six deep learning and nine machine learning techniques. Suitability of multimodel ensembles (MMEs) including arithmetic mean of all the models (Ens1), average of the best three performing models (Ens2), and weighted mean of outputs from all the 15 models was investigated. A total of nine “goodness-of-fit” measures (such as distance correlation (rd) and Taylor skill score) were used to rank the models. The two best deep learning methods were the nonlinear autoregressive model with exogenous input (NARX) ( r d = 0.51 ) and feedforward backpropagation (FFB) ( r d = 0.61 ). The two best machine learning algorithms included random forests (RF) ( r d = 0.64 ) and Gaussian process regression (GPR) ( r d = 0.59 ). Eventually, Ens2 was obtained using RF, FFB, and GPR. Ens2 performed better than Ens1 and Ens3. The amount of variance explained by individual models and MMEs was over the ranges of 13–66% and 51–74%, respectively. Ens2 explained 74% of the total variance in observed residual chlorine. Remarkably, the appropriateness of the MMEs depends on the approach for combining model outputs, and the number of models considered. This study demonstrated the acceptability of statistical MMEs in predicting chlorine residual concentration in drinking water.

Adoption of Artificial Intelligence in the Ugandan Health Sector: a review of Literature

“…The algorithm achieved an accuracy of 98.6%. Another study by Onyutha et al [24] utilized six deep learning and nine machine learning techniques to develop algorithms that can predict residual chlorine in drinking water. Their bestperforming model managed to explain 74% of the total variance in observed residual chlorine.…”

“…Implementing an AI-based system does not come cheap as many associated activities and requirements are costly or even hard to obtain especially in developing countries like Uganda. Authors of the reviewed papers highlighted the financial hardships faced in acquiring and annotating datasets [17,24], hardware and computing resources [24,43,49], and system maintenance and upgrading [29,42]. Modern AI-driven systems leverage deep learning neural networks that require very high computing hardware compared to traditional machine learning algorithms.…”

“…Most of the AI systems reviewed in this study focused on training, validation, and testing of the models; they were never used or deployed beyond an academic setting [24,29,42,43,49]. This shows the wide gap between AI systems development and deployment in healthcare practice.…”