Conference Presentations by Temitope O L U F U N M I Atoyebi
Comparison of Multinomial Naive Bayes (MNB), Gaussian Naive Bayes (GNB) and Random Forest (RF) Algorithm in Malaria Disease Diagnosis
IEEE Xplore 2024 International Conference on Science, Engineering and Business for Driving Sustainable Development Goals (SEB4SDG), 2024
Abstract:
Malaria, caused by Plasmodium parasites and transmitted by infected female Anopheles mo... more Abstract:
Malaria, caused by Plasmodium parasites and transmitted by infected female Anopheles mosquitoes, is still a major worldwide health problem. Millions of people are affected by the disease, which causes life-threatening symptoms and can be deadly in tropical locations. Effective malaria prevention and treatment need early and precise identification. As a result, this study investigates the ability of three machine learning algorithms including Multinomial Naïve Bayes (MNB), Gaussian Naive Bayes (GNB) and Random Forest (RF), in improving performance of malaria diagnosis. The MNB, GNB and RF classifiers are trained to identify malaria cases using a diversified dataset with factors such as "pregnancies," "treated net," "water breeding," "infected mosquitoes," and "age." The dataset consists of patients' health records from General Hospitals in Bwari Area Council and Abuja Municipal Area Council, Abuja, Nigeria. The dataset is gotten from year 2017 to 2021 and it includes both numerical and categorical variables. A 10-fold cross-validation procedure is used to evaluate the model. The experimental findings show that the Random Forest and Gaussian naïve Bayes algorithms performed well with 100% accuracy while the Multinomial Naive Bayes classifier performed better with 97% accuracy. The findings from this study can guide the selection of appropriate machine learning algorithms for real-world malaria diagnosis tasks based on their performance and also tend to addressing paradigm shift from laboratory to the applied Computing field.
Uploads
Conference Presentations by Temitope O L U F U N M I Atoyebi
Malaria, caused by Plasmodium parasites and transmitted by infected female Anopheles mosquitoes, is still a major worldwide health problem. Millions of people are affected by the disease, which causes life-threatening symptoms and can be deadly in tropical locations. Effective malaria prevention and treatment need early and precise identification. As a result, this study investigates the ability of three machine learning algorithms including Multinomial Naïve Bayes (MNB), Gaussian Naive Bayes (GNB) and Random Forest (RF), in improving performance of malaria diagnosis. The MNB, GNB and RF classifiers are trained to identify malaria cases using a diversified dataset with factors such as "pregnancies," "treated net," "water breeding," "infected mosquitoes," and "age." The dataset consists of patients' health records from General Hospitals in Bwari Area Council and Abuja Municipal Area Council, Abuja, Nigeria. The dataset is gotten from year 2017 to 2021 and it includes both numerical and categorical variables. A 10-fold cross-validation procedure is used to evaluate the model. The experimental findings show that the Random Forest and Gaussian naïve Bayes algorithms performed well with 100% accuracy while the Multinomial Naive Bayes classifier performed better with 97% accuracy. The findings from this study can guide the selection of appropriate machine learning algorithms for real-world malaria diagnosis tasks based on their performance and also tend to addressing paradigm shift from laboratory to the applied Computing field.