Efficient artificial intelligence forecasting models for COVID-19 outbreak in Russia and Brazil

Area of Science:

• Epidemiology
• Computational Intelligence
• Public Health

Background:

• COVID-19, a pandemic declared by WHO, impacts multiple human systems and necessitates effective forecasting for policy development.
• Accurate prediction of COVID-19 cases, especially in hotspots, is vital for global health security and resource allocation.
• Existing forecasting models require enhancement to address the rapid spread and multifaceted impact of the virus.

Purpose of the Study:

• To propose a novel short-term forecasting model for COVID-19 cases.
• To enhance the adaptive neuro-fuzzy inference system (ANFIS) using an improved optimization algorithm.
• To evaluate the performance of the proposed model against existing artificial intelligence methods.

Main Methods:

• Development of a chaotic marine predators algorithm (CMPA) to optimize the ANFIS model.
• Comparison of the proposed chaotic marine predators algorithm-enhanced ANFIS (CMPA-ANFIS) with original ANFIS, ANFIS with marine predators algorithm (MPA-ANFIS), and ANFIS with particle swarm optimization (PSO-ANFIS).
• Statistical assessment of forecasting accuracy using various criteria.

Main Results:

• The proposed CMPA-ANFIS model demonstrated significantly superior forecasting accuracy compared to all other evaluated models.
• The chaotic MPA effectively addressed the shortcomings of the original ANFIS, leading to improved predictive performance.
• Statistical analysis confirmed the robustness and effectiveness of the CMPA-ANFIS for COVID-19 case forecasting.

Conclusions:

• The CMPA-ANFIS model offers a promising approach for accurate short-term COVID-19 forecasting.
• Optimizing ANFIS with advanced algorithms like CMPA is crucial for improving epidemiological predictions.
• This study provides a valuable tool for policymakers to manage the COVID-19 pandemic effectively.