Which Machine Learning method for outbreaks predictions?
- 27 January 2021
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 825-828
Abstract
African swine fever (ASF), dengue fever, influenza, norovirus disease are names of outbreaks that have taken some parts of the world by surprise and caused several deaths. The advent of coronavirus disease at the end of 2019 is clear proof that the issue of outbreaks is more topical than ever. The Machine Learning (ML) comes to the rescue of medicine by offering predictive analysis methods to predict the reappearance of outbreaks and thus help the healthcare system and decisionmakers to take the necessary measures in advance. But which ML method should be used to predict an outbreak? Given the large number of ML methods, it is sometimes difficult to know which method to use. This is why we conducted this study. We looked at the above-mentioned outbreaks and identified the following ML methods as the most commonly used: LR, C4.5, NB, SVM, ANN and RF. Of these six methods, the random forest method (RF) stands out from the others because it was used in all four outbreaks considered, has the best accuracy for two outbreaks (ASF and dengue fever) and has been designated by several independent researchers as the most suitable ML method for outbreak prediction.Keywords
This publication has 17 references indexed in Scilit:
- Mobile Data Collection Using Open Data Kit2019
- Prediction for global African swine fever outbreaks based on a combination of random forest algorithms and meteorological dataTransboundary and Emerging Diseases, 2019
- Comparative evaluation of time series models for predicting influenza outbreaks: application of influenza-like illness data from sentinel sites of healthcare centers in IranBMC Research Notes, 2019
- Deep Learning for Epidemiological Predictions2018
- Supervised forecasting of the range expansion of novel non‐indigenous organisms: Alien pest organisms and the 2009 H1N1 flu pandemicGlobal Ecology and Biogeography, 2018
- Development of artificial intelligence approach to forecasting oyster norovirus outbreaks along Gulf of Mexico coastEnvironment International, 2018
- Long-term predictors of dengue outbreaks in Bangladesh: A data mining approachInfectious Disease Modelling, 2018
- Development of genetic programming-based model for predicting oyster norovirus outbreak risksWater Research, 2018
- The global distribution and burden of dengueNature, 2013
- Decision Tree Algorithms Predict the Diagnosis and Outcome of Dengue Fever in the Early Phase of IllnessPLOS Neglected Tropical Diseases, 2008