Hybrid grey assisted whale optimization based machine learning for the COVID-19 prediction

Area of Science:

• Computational biology
• Epidemiology
• Artificial Intelligence

Background:

• The COVID-19 pandemic has caused significant socio-economic disruption and mortality.
• Rapid spread and lack of a definitive vaccine necessitate accurate prediction models.
• Existing prediction methods may require improvement for accuracy and efficiency.

Purpose of the Study:

• To develop and evaluate a machine learning scheme for predicting COVID-19 positive, negative, and deceased instances.
• To enhance prediction accuracy using feature selection and a robust classification model.
• To assess the model's performance against established metrics using real-world data.

Main Methods:

• Data preprocessing involved removing redundant and missing values.
• Feature selection was performed using the hybrid grey assisted whale optimization algorithm (H-GAWOA).
• The adaptive network-based fuzzy inference system (ANFIS) classifier was employed for outcome prediction.

Main Results:

• The proposed H-GAWOA-ANFIS model demonstrated superior performance with very low Mean Squared Error (MSE) values across all predictions.
• Specifically, the model achieved an MSE of 0.00 for predicting deceased COVID-19 cases.
• Performance metrics including MSE, Root Mean Squared Error (RMSE), Mean Absolute Percentage Error (MAPE), and R-squared were analyzed.

Conclusions:

• The H-GAWOA-ANFIS model significantly outperforms other approaches in predicting COVID-19 outcomes.
• The proposed machine learning scheme offers a promising tool for epidemiological analysis and public health management.
• Accurate prediction of COVID-19 cases, including mortality, can aid in resource allocation and intervention strategies.