Analysis of malaria and climate in Damaturu City of Nigeria using predictive supervised learning

Malaria has always been a substantial health-related challenge in Nigeria due to climate variabilities emerging as a critical driver of its incidence. This study investigates the underlying nexus, connecting climatic factors, and malaria prevalence in Damaturu, Nigeria, using supervised machine learning techniques on a decade-long (2014–2023) data set from the Epidemiologic Data Repository and climate data from the Nigeria Meteorological Agency (NiMet). Results revealed a moderate relationship connecting rainfall and malaria occurrence, showing that increased rainfall creates conducive conditions for mosquito breeding and pathogens. Conversely, there is a weak relationship connecting malaria cases and temperature. This implies that higher temperatures may disrupt mosquito survival. Predicting malaria cases was achieved by leveraging LinearRegressor (RMSE score = 1989.37), DecisionTreeRegressor (1965.08), LassoCV (Lasso Cross Validation, 1809.65), and evaluated using root mean squared error (RMSE). GridSearchCV (1630.51) was also deployed in running cross-validation, after which Ensemble techniques like Voting (1622.03), Stacking (1628.30), and RandomForest (1773.07) were also employed in reducing predictive error. The Voting model was observed to have exhibited the least error (RMSE value) when compared with the actual malaria cases from the predicted, RandomForestRegressor's GridSearchCV model with an error score of 1630 was finally adopted because it was observed to be most suitable in modeling the non-linear patterns among the features, in minimizing predictive error, in minimizing overfitting or underfitting, and the most effective in the prediction of malaria incidence in Damaturu. This underscores the importance of seasonally informed climate intervention. The findings further show the importance of community-based interventions, especially during high-rainfall periods, as well as ensuring that malaria control programs consider climatic factors and variabilities in localized areas. The implication of this study is to reduce the malaria burden and its consequent socio-economic variability.