Short-Term Forecasting of Monkeypox Cases Using a Novel Filtering and Combining Technique.

Area of Science:

• Public Health Surveillance
• Epidemiology
• Computational Biology

Background:

• Modern healthcare surveillance systems increasingly rely on advanced technologies like artificial intelligence (AI), machine learning (ML), and big data analytics.
• Accurate monitoring of infectious diseases, such as monkeypox (MPX), is critical for timely public health responses.
• Publicly available datasets on confirmed cases offer potential for developing predictive models.

Purpose of the Study:

• To propose a novel filtering and combination technique for accurate short-term forecasting of infected monkeypox cases.
• To enhance the prediction accuracy of early-stage confirmed monkeypox cases using machine learning models.
• To provide a reliable forecasting tool for understanding disease spread and associated risks.

Main Methods:

• Time series analysis involving filtering cumulative confirmed cases into long-term trend and residual subseries using proposed and benchmark filters.
• Prediction of filtered subseries using five standard machine learning models and their combinations.
• Direct combination of individual forecasting models for one-day-ahead predictions of newly infected cases, validated by error metrics and statistical tests.

Main Results:

• The proposed filtering and combination technique demonstrated significant efficiency and accuracy in forecasting monkeypox cases.
• Experimental results confirmed the superiority of the proposed methodology compared to benchmark time series and machine learning models.
• A fourteen-day forecast was achieved using the best combination model, indicating robust predictive capability.

Conclusions:

• The developed forecasting methodology effectively predicts short-term monkeypox case numbers, aiding in understanding disease dynamics.
• Accurate forecasting supports risk assessment, enabling proactive public health interventions to prevent further spread.
• The approach facilitates timely and effective treatment strategies through improved disease monitoring.