Influence of Environmental Variables on COVID-19 Pandemic in the Kano State of Nigeria: An Artificial Intelligence-Based Approach

Abstract

This study investigates the influence environmental variables on COVID-19 pandemic in Kano State, Nigeria through the artificial intelligence-based modelling methods. A descriptive correlational quantitative design was a design based on secondary data collected in a period of four months, between 1st December 2020 and 31^st March 2021, with a total of 121 daily observations. The Nigerian Centre for Disease Control (NCDC) and Kano State Ministry of Health provided the COVID-19 epidemiological data (new cases, recovery cases, and death cases) whereas the Nigerian Meteorological Agency (NiMet) and AccuWeather sources were used to retrieve the environmental data (temperature, humidity, and wind speed). The methods of artificial intelligence were used in the MATLAB with three individual models Multiple Linear Regression (MLR), Artificial Neural Network (ANN), and Adaptive Neuro-Fuzzy Inference System (ANFIS). The coefficient of determination (R²), correlation coefficient (R) as well as root mean square error (RMSE) were used to assess model performance. In addition, hybrid models (MLR-ANN and MLR-ANFIS) were also employed to increase predictive performance. The findings showed that the hybrid models outperformed the individual models as they showed better R2 and reduced RMSE during training and testing. The results present the possibility of hybrid artificial intelligence methods in enhancing predictive modeling of COVID-19 dynamics on environmental factors.