Forecasting of Monkeypox New Cases in North and South America for 2023: A Comparative Analysis Using Machine Learning

Abstract

The recent spread of monkeypox throughout the world highlights the critical need for trustworthy forecasting models to guide public health actions. This study evaluates several machine learning techniques for forecasting new monkeypox cases in North and South America in 2023. We applied five models Decision Tree (DT), Artificial Neural Network (ANN), K-Nearest Neighbors (KNN), XGBoost, and Random Forest (RF) and evaluated them using four key error metrics: Mean Squared Error (MSE), Root Mean Squared Error (RMSE), Mean Absolute Error (MAE), and Mean Absolute Percentage Error (MAPE). For South America, the Decision Tree model produced the most accurate forecasts, with the lowest MAE (10.37), RMSE (13.28), MSE (176.25), and MAPE (28.62%). Meanwhile, XGBoost proved to be the best-performing model for North America, with the lowest MAE (17.06), RMSE (19.62), MSE (384.79), and MAPE (25.34%). Our results highlight the significance of region-specific modeling strategies since various machine learning methods performed better in multiple environments. This research lays the groundwork for creating more precise forecasting instruments that will aid in halting the monkeypox epidemic and improve response and readiness throughout North and South America.

Introduction

Monkeypox virus (MPXV) was first identified in humans in 1970, mainly affecting Central and West Africa. Unlike SARS-CoV-2, MPXV spreads primarily through close contact after symptom onset, with no strong evidence of asymptomatic or long-range airborne transmission. Since the eradication of smallpox, monkeypox has emerged as a significant zoonotic threat with a fatality rate of ~11%. Traditionally endemic, the virus has caused widespread outbreaks in non-endemic regions since May 2022, particularly in North and South America, driven by the West African clade. Human-to-human transmission occurs mainly via respiratory droplets, bodily fluids, or contact with mucous membranes and lesions, while squirrels and striped grass mice serve as primary reservoirs.

The 2022–2023 outbreaks saw simultaneous cases across multiple countries, challenging healthcare systems. Effective surveillance and predictive models are needed for epidemic control. This study uses machine learning (ML) techniques to forecast monkeypox trends in North and South America. The models applied include Artificial Neural Networks (ANN), Multilayer Perceptron (MLP), Decision Trees, Random Forest, k-Nearest Neighbor (KNN), and XGBoost, trained on daily new cases from January to December 2023. Data were split into 75% training and 25% testing sets, and performance was evaluated using MAE, MSE, RMSE, and MAPE metrics.

Results show that both continents experienced a similar trend: a peak in daily new cases in January 2023, a sharp decline in February–March, stable low transmission from April to September, and a late-year increase peaking in November before a small December decline. North America had higher overall case numbers than South America. The ML models aim to improve outbreak forecasting, enabling better-targeted public health interventions.

Conclusion

This study examined various machine learning models, including ANN, KNN, Decision Tree, Random Forest, and XGBoost, to forecast new cases of monkeypox in North and South America for 2023. We found that XGBoost delivered the best performance in North America, whereas the Decision Tree model yielded the most accurate predictions in South America. The variation in model performance between North and South America highlights how regional factors can influence the choice of the best forecasting model. In South America, the Decision Tree model worked well due to its straightforward nature, while in North America, XGBoost’ s ability to capture more complex trends made it the better option. These results show how important it is to adjust forecasting models based on the unique characteristics of each region. Although the results are promising, there are some limitations to consider. The small amount of case data, the lack of environmental and demographic factors, and the sensitivity to hyperparameters highlight the need for further improvements. By choosing the right models, health officials in North and South America can make more accurate predictions about monkeypox cases and act quickly. In the future, combining different models or trying out hybrid approaches could help improve the accuracy of forecasts across various regions.

References

[1] Nuzzo JB, Borio LL, Gostin LO. The WHO Declaration of Monkeypox as a Global Public Health Emergency. JAMA. 2022 Aug 16;328(7):615–7. [2] Heymann DL, Szczeniowski M, Esteves K. Re-emergence of monkeypox in Africa: a review of the past six years. Br Med Bull. 1998;54(3):693–702. [3] Durski KN. Emergence of Monkeypox — West and Central Africa, 1970–2017. MMWR Morb Mortal Wkly Rep [Internet]. 2018 [cited 2024 Oct 14];67. Available from: https://www.cdc.gov/mmwr/volumes/67/wr/mm6710a5.htm [4] Bankuru SV, Kossol S, Hou W, Mahmoudi P, Rychtá? J, Taylor D. A game-theoretic model of Monkeypox to assess vaccination strategies. PeerJ. 2020 Jun 22;8:e9272. [5] Wei W, Wang G, Tao X, Luo Q, Chen L, Bao X, et al. Time series prediction for the epidemic trends of monkeypox using the ARIMA, exponential smoothing, GM (1, 1) and LSTM deep learning methods. J Gen Virol [Internet]. 2023 Apr 6 [cited 2024 Oct 14];104(4). Available from: https://www.microbiologyresearch.org/content/journal/jgv/10.1099/jgv.0.001839 [6] Kmiec D, Kirchhoff F. Monkeypox: a new threat? Int J Mol Sci. 2022;23(14):7866. [7] Karem KL, Reynolds M, Hughes C, Braden Z, Nigam P, Crotty S, et al. Monkeypox-induced immunity and failure of childhood smallpox vaccination to provide complete protection. Clin Vaccine Immunol CVI. 2007 Oct;14(10):1318–27. [8] Zhang WD, Zu ZH, Xu Q, Xu ZJ, Liu JJ, Zheng T. Optimized strategy for the control and prevention of newly emerging influenza revealed by the spread dynamics model. PloS One. 2014;9(1):e84694. [9] J Z, Ml L, Zf L, Mw S, Yn X, Fp J. Transmission patterns of COVID-19 in the mainland of China and the efficacy of different control strategies: a data- and model-driven study. Infect Dis Poverty [Internet]. 2020 Jul 6 [cited 2024 Oct 14];9(1). Available from: https://pubmed.ncbi.nlm.nih.gov/32631426/ [10] Jr W, Sn I. Monkeypox virus and insights into its immunomodulatory proteins. Immunol Rev [Internet]. 2008 Oct [cited 2024 Oct 14];225. Available from: https://pubmed.ncbi.nlm.nih.gov/18837778/ [11] Ngbolua K te N, Ngambika GK, Mbembo-wa-Mbembo B, Séraphin KP, Fabrice KK, Bongo GN, et al. First Report on Three Cases of Monkey pox in Nord Ubangi Province (Democratic Republic of the Congo). Br Int Exact Sci BIoEx J. 2020 Jan 9;2(1):120–5. [12] Status of human monkeypox: clinical disease, epidemiology and research - ScienceDirect [Internet]. [cited 2024 Oct 14]. Available from: https://www.sciencedirect.com/science/article/pii/S0264410X1100524X [13] 2022–2023 mpox outbreak in South America. In: Wikipedia [Internet]. 2024 [cited 2024 Oct 14]. Available from: https://en.wikipedia.org/w/index.php?title=2022%E2%80%932023_mpox_outbreak_in_South_America&oldid=1236886127 [14] 2022–2023 mpox outbreak in North America. In: Wikipedia [Internet]. 2024 [cited 2024 Oct 14]. Available from: https://en.wikipedia.org/w/index.php?title=2022%E2%80%932023_mpox_outbreak_in_North_America&oldid=1236886126 [15] Dada EG, Oyewola DO, Joseph SB, Emebo O, Oluwagbemi OO. Ensemble Machine Learning for Monkeypox Transmission Time Series Forecasting. Appl Sci. 2022 Jan;12(23):12128. [16] Ncube B, Dzikiti M, Nyoni A, Ncube M, Ndlovu BM, Dube S. Effectiveness of Machine Learning algorithms in predicting Monkey Pox (Mpox): A Systematic Literature Review. [17] Priyadarshini I, Mohanty P, Kumar R, Taniar D. Monkeypox Outbreak Analysis: An Extensive Study Using Machine Learning Models and Time Series Analysis. Computers. 2023 Feb 7;12(2):36. [18] Mathieu E, Spooner F, Dattani S, Ritchie H, Roser M. Mpox. Our World Data [Internet]. 2024 Mar 22 [cited 2024 Oct 14]; Available from: https://ourworldindata.org/mpox [19] Databricks [Internet]. 2022 [cited 2024 Oct 15]. What are Machine Learning Models? Available from: https://www.databricks.com/glossary/machine-learning-models [20] Forbytes [Internet]. 2023 [cited 2024 Oct 15]. Top 6 Machine Learning Techniques for Predictive Modeling. Available from: https://forbytes.com/blog/main-machine-learning-techniques/ [21] Neural network (machine learning). In: Wikipedia [Internet]. 2024 [cited 2024 Oct 15]. Available from: https://en.wikipedia.org/w/index.php?title=Neural_network_(machine_learning)&oldid=1250750800 [22] Artificial Neural Network - an overview | ScienceDirect Topics [Internet]. [cited 2024 Oct 15]. Available from: https://www.sciencedirect.com/topics/neuroscience/artificial-neural-network [23] Yang Z, Yang Z (2014). Comprehensive Biomedical Physics. Karolinska Institutet, Stockholm, Sweden: Elsevier. p. 1. ISBN 978-0-444-53633-4. Archived from the original on 28 July 2022. Retrieved 28 July 2022 - Google Search [Internet]. [cited 2024 Oct 15]. Available from: https://www.google.com/search?client=firefox-b-d&sca_esv=abee16815fc6fe5b&q=Yang+Z,+Yang+Z+(2014).+Comprehensive+Biomedical+Physics.+Karolinska+Institutet,+Stockholm,+Sweden:+Elsevier.+p.+1.+ISBN+978-0-444-53633-4.+Archived+from+the+original+on+28+July+2022.+Retrieved+28+July+2022&sa=X&ved=2ahUKEwix1qi6k5CJAxW6fGwGHQEYNoAQ7xYoAHoECAoQAQ&biw=1366&bih=643&dpr=1 [24] (01) The Nature of Statistical Learning Theory.pdf [Internet]. [cited 2024 Oct 15]. Available from: http://39.108.104.114/res/(01)%20The%20Nature%20of%20Statistical%20Learning%20Theory.pdf [25] Hornik K, Stinchcombe M, White H. Multilayer feedforward networks are universal approximators. Neural Netw. 1989 Jan 1;2(5):359–66. [26] www.javatpoint.com [Internet]. [cited 2024 Oct 15]. Decision Tree Algorithm in Machine Learning - Javatpoint. Available from: https://www.javatpoint.com/machine-learning-decision-tree-classification-algorithm [27] scikit-learn [Internet]. [cited 2024 Oct 15]. 1.10. Decision Trees. Available from: https://scikit-learn/stable/modules/tree.html [28] Ben-Haim Y, Tom-Tov E. A Streaming Parallel Decision Tree Algorithm. [29] Abdulhafedh A. Comparison between Common Statistical Modeling Techniques Used in Research, Including: Discriminant Analysis vs Logistic Regression, Ridge Regression vs LASSO, and Decision Tree vs Random Forest. OALib. 2022;09(02):1–19. [30] Tin Kam Ho. The random subspace method for constructing decision forests. IEEE Trans Pattern Anal Mach Intell. 1998 Aug;20(8):832–44. [31] Random Forest Classification with Scikit-Learn [Internet]. [cited 2024 Oct 15]. Available from: https://www.datacamp.com/tutorial/random-forests-classifier-python [32] Oyewola D, Hakimi D, Adeboye K, Shehu MD. Using Five Machine Learning for Breast Cancer Biopsy Predictions Based on Mammographic Diagnosis. Int J Eng Technol IJET. 2017 Apr 20;2(4):142–5. [33] Nair A. AIM. 2019 [cited 2024 Oct 15]. MLP Classifier - A Beginner’s Guide To SKLearn MLP Classifier. Available from: https://analyticsindiamag.com/ai-mysteries/a-beginners-guide-to-scikit-learns-mlpclassifier/ [34] Classification Using Sklearn Multi-layer Perceptron - GeeksforGeeks [Internet]. [cited 2024 Oct 15]. Available from: https://www.geeksforgeeks.org/classification-using-sklearn-multi-layer-perceptron/ [35] Deepchecks [Internet]. [cited 2024 Oct 15]. What is Multilayer Perceptron. Available from: https://www.deepchecks.com/glossary/multilayer-perceptron/ [36] Dutta S. Multi-Layer Perceptron and Backpropagation: A Deep Dive [Internet]. Medium. 2024 [cited 2024 Oct 15]. Available from: https://medium.com/@sanjay_dutta/multi-layer-perceptron-and-backpropagation-a-deep-dive-8438cc8bcae6 [37] What is k-Nearest Neighbor (kNN)? | A Comprehensive k-Nearest Neighbor Guide [Internet]. [cited 2024 Oct 15]. Available from: https://www.elastic.co/what-is/knn [38] GeeksforGeeks [Internet]. 2017 [cited 2024 Oct 15]. K-Nearest Neighbor(KNN) Algorithm. Available from: https://www.geeksforgeeks.org/k-nearest-neighbours/ [39] What is the k-nearest neighbors algorithm? | IBM [Internet]. 2021 [cited 2024 Oct 15]. Available from: https://www.ibm.com/topics/knn [40] xgboost algorithm - Google Search [Internet]. [cited 2024 Oct 16]. Available from: https://www.google.com/search?q=xgboost+algorithm&client=firefox-b-d&sca_esv=28f8fab5923385a6&ei=jK8OZ-OTB_rZseMP0vn0qQc&oq=XGBoost&gs_lp=Egxnd3Mtd2l6LXNlcnAiB1hHQm9vc3QqAggBMgsQABiABBiRAhiKBTILEAAYgAQYkQIYigUyChAAGIAEGEMYigUyChAAGIAEGEMYigUyChAAGIAEGEMYigUyChAAGIAEGEMYigUyBRAAGIAEMgUQABiABDIFEAAYgAQyBRAAGIAESJFRUIAVWOA3cAN4AZABAJgBuwGgAbwIqgEDMC43uAEByAEA-AEBmAIKoALJCqgCCsICChAAGLADGNYEGEfCAhQQABiABBiRAhi0AhiKBRjqAtgBAcICFhAAGAMYtAIY5QIY6gIYjAMYjwHYAQHCAgsQABiABBixAxiDAcICCxAuGIAEGLEDGIMBwgIOEAAYgAQYsQMYgwEYigXCAhEQLhiABBixAxjRAxiDARjHAcICEBAuGIAEGNEDGEMYxwEYigXCAggQLhiABBixA5gDDOIDBRIBMSBAiAYBkAYIugYECAEYB5IHBTMuNi4xoAfgMA&sclient=gws-wiz-serp [41] Simplilearn.com [Internet]. 2022 [cited 2024 Oct 16]. What is XGBoost? An Introduction to XGBoost Algorithm in Machine Learning | Simplilearn. Available from: https://www.simplilearn.com/what-is-xgboost-algorithm-in-machine-learning-article [42] GeeksforGeeks [Internet]. 2021 [cited 2024 Oct 16]. XGBoost. Available from: https://www.geeksforgeeks.org/xgboost/ [43] Torres LF. XGBoost: The King of Machine Learning Algorithms [Internet]. LatinXinAI. 2023 [cited 2024 Oct 16]. Available from: https://medium.com/latinxinai/xgboost-the-king-of-machine-learning-algorithms-6b5c0d4acd87 [44] Bro?yna J, Mentel G, Szetela B, Strielkowski W. Multi-Seasonality in the TBATS Model Using Demand for Electric Energy as a Case Study. Econ Comput Econ Cybern Stud Res Acad Econ Stud. 2018 Mar 20;52:229–46. [45] Bengesi S, Oladunni T, Olusegun R, Audu H. A Machine Learning-Sentiment Analysis on Monkeypox Outbreak: An Extensive Dataset to Show the Polarity of Public Opinion From Twitter Tweets. IEEE Access. 2023;11:11811–26. [46] Qureshi M, Khan S, Bantan RAR, Daniyal M, Elgarhy M, Marzo RR, et al. Modeling and Forecasting Monkeypox Cases Using Stochastic Models. J Clin Med. 2022 Jan;11(21):6555. [47] Detection of Monkeypox Cases Based on Symptoms Using XGBoost and Shapley Additive Explanations Methods [Internet]. [cited 2024 Oct 17]. Available from: https://www.mdpi.com/2075-4418/13/14/2391 [48] Akinola SO, Wang QG, Olukanmi P, Marwala T. Early Prediction of Monkeypox Virus Outbreak Using Machine Learning. IETI Trans Data Anal Forecast ITDAF. 2023 Jul 6;1(2):14–29.

Copyright

Copyright © 2025 Sharmin Akther, Amrin Binte Ahmed, Sanjida Tasnim. This is an open access article distributed under the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.