Core Concepts
The author presents an improved forecasting model using the PSO-RDV framework to enhance the accuracy of Artificial Neural Network predictions for HIV cases in the Philippines.
Abstract
The study focuses on optimizing forecasting models for HIV incidence in the Philippines. It introduces a novel approach combining Particle Swarm Optimization (PSO) with Random Descending Velocity Inertia Weight (RDV IW) technique to improve convergence and accuracy. The research highlights significant improvements in position error, computational time, and convergence compared to traditional models. The proposed model demonstrates superior performance in terms of accuracy metrics, providing valuable insights for future forecasting systems.
Stats
Simulation results revealed a 6.36% improvement in position error and 11.75% improvement in computational time.
Computed p-values for NRMSE (0.04889174), MAE (0.02829063), MAPE (0.02226053), WAPE (0.01701545), and R2 (0.00000021) indicated significant results.
The optimal combination of RDV IW control parameters was found to be alpha = 0.4 and alpha_dump = 0.9.
Quotes
"The proposed forecasting model exhibited a 6.36% improvement in position error."
"The computed p-values indicated significant results in terms of accuracy performance."