Open AccessPrediction of the effects of environmental factors towards COVID-19 outbreak using AI-based models
Author(s) -
Khalid Mahmoud,
Hatice Bebiş,
A. G. Usman,
Armend Salihu,
Muhammad Sani Gaya,
Umar Farouk Dalhat,
Rabiu Aliyu Abdulkadir,
Mustapha Bala Jibril,
S.I. Abba
Publication year - 2021
Publication title -
iaes international journal of artificial intelligence
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.341
H-Index - 7
eISSN - 2252-8938
pISSN - 2089-4872
DOI - 10.11591/ijai.v10.i1.pp35-42
Subject(s) - adaptive neuro fuzzy inference system , mean squared error , correlation coefficient , coefficient of determination , covid-19 , wind speed , calibration , statistics , artificial neural network , spearman's rank correlation coefficient , computer science , linear regression , mathematics , environmental science , artificial intelligence , fuzzy logic , meteorology , fuzzy control system , geography , medicine , disease , pathology , infectious disease (medical specialty)
The need for elucidating the effects of environmental factors in the determination of the novel corona virus (COVID-19) is very vital. This study is a methodological study to compare three different test models (1. Artificial neural networks (ANN), 2. Adaptive neuro fuzzy inference system (ANFIS), 3. A linear classical model (MLR)) used to determine the relationship between COVID-19 spread and environmental factors (temperature, humidity and wind). These data were obtained from the studies (Pirouz, Haghshenas, Haghshenas, & Piro, 2020) with confirmed COVID-19 patients in Wuhan, China, using temperature, humidity and wind as the independent variables. The measured and the predicted results were checked based on three different performance indices; Root mean square error (RMSE), determination coefficient (R2) and correlation coefficient (R). The results showed that ANFIS and ANN are more promising over the classical MLR models having an average R-values of 0.90 in both calibration and verification stages. The findings indicated that ANFIS outperformed MLR and ANN. In addition, their performance skills boosted up to 25% and 9% respectively based on the determination coefficient for the prediction of confirmed COVID-19 cases in Wuhan city of China. Overall, the results depict the reliability and ability of AI-based models (ANFIS and ANN) for the simulation of COVID-19 using the effects of various environmental variables.