Machine Learning Techniques for Predicting COVID-19 Outbreak Trends: A Multi-Algorithm Approach

COVID-19 pandemic has exposed vulnerabilities in public health systems globally, highlighting the need for precise predictive methodologies to guide interventions and resource allocation. This study employs machine learning techniques to estimate epidemic trends, including daily cases, recoveries, and deaths. Using global COVID-19 data, algorithms such as LR, SVM, RF, PCA, and LDA were applied. Among these, the RF model emerged as the most effective for predicting active cases, achieving an accuracy of 96% and a precision of 95%. Key predictive factors, including daily new cases and growth rate, were identified. The integration of dimensionality reduction and feature engineering alongside machine learning models further enhanced predictive performance. These findings offer actionable insights for healthcare systems and policymakers, enabling proactive approaches that can mitigate the impact of pandemics and reduce reliance on reactive strategies such as lockdowns. This research contributes to pandemic preparedness by providing data-driven tools that improve public health outcomes.

After attending this session, participants will be able to analyze how different machine learning algorithms (LR, SVM, RF, PCA, LDA) can be applied to outbreak data and evaluate their comparative performance in predicting epidemic trends.

After attending this session, participants will be able to apply predictive modeling concepts to real-world public health challenges by using machine learning approaches to guide decision-making, resource allocation, and proactive pandemic preparedness.

After attending this session, participants will be able to design data-driven strategies that integrate feature engineering and dimensionality reduction to improve accuracy and reliability in forecasting infectious disease outbreaks.