Predicting malaria outbreak in The Gambia using machine learning techniques

Abstract

Malaria is the most common cause of death among the parasitic diseases. Malaria continues to pose a growing threat to the public health and economic growth of nations in the tropical and subtropical parts of the world. This study aims to address this challenge by developing a predictive model for malaria outbreaks in each district of The Gambia, leveraging historical meteorological data. To achieve this objective, we employ and compare the performance of eight machine learning algorithms, including C5.0 decision trees, artificial neural networks, k-nearest neighbors, support vector machines with linear and radial kernels, logistic regression, extreme gradient boosting, and random forests. The models are evaluated using 10-fold cross-validation during the training phase, repeated five times to ensure robust validation. Our findings reveal that extreme gradient boosting and decision trees exhibit the highest prediction accuracy on the testing set, achieving 93.3% accuracy, followed closely by random forests with 91.5% accuracy. In contrast, the support vector machine with a linear kernel performs less favorably, showing a prediction accuracy of 84.8% and underperforming in specificity analysis. Notably, the integration of both climatic and non-climatic features proves to be a crucial factor in accurately predicting malaria outbreaks in The Gambia.

Fig 1. The map of the Gambia showing its five main meteorological stations across different regions.

Fig 2. Distribution of climatic variables, malaria cases and deaths: Malaria cases, deaths and rainfall are highly skewed to the right.

None of the features appeared to be distributed normally.

Fig 3. Proportion of outbreak in both the training and testing data.

Fig 4. The Pearson’s correlation among all the climatic variables used.

Fig 5. Variation of the climatic variables, total malaria cases and deaths in each month over the years.

In the upper panels of the first row, we can observe the number of malaria cases and deaths in thousands, although not used as predictors. Each line represents a different year, indicated by the various colors. The bottom four panels show the climate variables and their variability. The boxplots depict each month of the year, while the colored dots represent the data points in the corresponding years. The minimum and maximum temperatures were measured in degrees Celsius (°C), the rainfall was measured in millimeters (mm), and the relative humidity was measured in percentages (%).

Fig 6. This plot shows the area under the Receiver Operating Characteristic (ROC), (AUC) of all the learning algorithms.

The green and the blue curve shows the ROC curve using the training and testing data respectively. XGBoost has the highest AUC value of 0.97. (a) C5.0 DT. (b) KNN. (c) Logistic Regression. (d) ANN. (e) Random Forest. (f) XGBoost. (g) svmLinear. (h) svmRadial.