Predictive modeling and intervention strategies for COVID-19 spread estimation using machine learning and computational intelligence techniques

This research examines the evolution of COVID-19 in Malaysia following vaccine implementation on February 24, 2021, employing sophisticated machine learning and artificial intelligence methodologies to forecast and address the pandemic efficiently. The study develops a hybrid computational intelligence model by employing the Particle Swarm Optimization algorithms (PSO), along with an Adaptive Neuro Fuzzy Inference System (ANFIS). This hybrid model provides accurate prediction of disease dynamics for a 10-day period with RMSE of (Formula presented.). Another deep learning model is developed by employing the predicting advanced Recurrent Neural Network (RNN) architectures such as Long Short-Term Memory (LSTM), which is characterized by its nonlinear prediction capabilities for the same period of prediction with RMSE of 0.0125. Both models demonstrate low error values when compared to actual data of COVID-19 within the Malaysian Community, making them dependable and effective tools for policymakers, even in the absence of a specific threshold for epidemic model precision.