Quantitative Forecasting of Malaria Parasite Using Machine Learning Models: MLR, ANN, ANFIS and Random Forest

Area of Science:

• Medical Informatics
• Computational Biology
• Public Health

Background:

• Malaria poses a significant socioeconomic challenge in Africa, with a high mortality rate.
• Accurate malaria prediction is crucial for effective public health interventions and resource allocation.

Purpose of the Study:

• To evaluate and compare the predictive performance of various machine learning models for malaria.
• To identify the most accurate model for malaria prediction to aid healthcare decision-making.

Main Methods:

• Utilized a dataset of 2207 patients, reduced to 15 criteria samples.
• Compared four machine learning models: Artificial Neural Networks (ANN), Adaptive Neuro-Fuzzy Inference Systems (ANFIS), Multi-Linear Regression (MLR), and Random Forest classifier.
• Employed assessment measures including R-squared (R²), Root Mean Square Error (RMSE), Mean Square Error (MSE), and adjusted correlation coefficient (R).

Main Results:

• Artificial Neural Networks (ANN) achieved the highest accuracy (99% R and R²) post-training.
• ANN outperformed ANFIS (97%), MLR (92%), and Random Forest (68%) in predictive power.
• Testing phase validated the superior performance of ANN; MLR models showed excellent accuracy with minimal errors in statistical forecasting.

Conclusions:

• ANN models offer superior accuracy for malaria prediction compared to ANFIS, MLR, and Random Forest.
• The findings provide valuable insights for improving malaria prediction models and supporting healthcare decision-making.
• Machine learning, particularly ANN, enhances the precision and efficiency of disease prediction, optimizing resource allocation in healthcare systems.