Scholarlink

<p>The COVID-19 pandemic has significantly altered global health, economies, and societal structures, making accurate daily case predictions essential for strategic planning and outbreak control. This study employs two machine learning models: a multi-layer perceptron (MLP) neural network and a bio-inspired sperm swarm optimization (SSO) algorithm hybridized with MLP (MLP-SSO), comparing their performance in forecasting daily COVID-19 deaths in Brazil, India, Russia, and the United States using time series data from January 20 to September 15, 2020. To ensure a realistic evaluation of the model’s forecasting capability, the data were chronologically split into training (first 70%) and test (remaining 30%) sets, preserving the temporal order of the time series. Additionally, the average mutual information method was utilized to determine optimal time lags, enabling the model to capture nonlinear relationships in the time series data. Model performance was evaluated using standard statistical metrics, including root mean square error (RMSE) and Pearson correlation coefficient (R). The results demonstrated that the MLP-SSO hybrid model outperformed the standalone MLP neural network across all four countries, significantly improving MLP performance. Specifically, the R increased from 0.8 to 0.94 in Brazil, from 0.78 to 0.92 in India, from 0.92 to 0.96 in Russia, and from 0.49 to 0.90 in the United States. Furthermore, the RMSE was reduced by 52.5% in Brazil, 9.3% in India, 84.0% in Russia, and 80.9% in the United States. These findings provide valuable insights for policymakers and health authorities in managing and controlling the COVID-19 pandemic.</p>

错误：搜索内容不能为空，请输入英文关键词
错误：关键词超出字数限制，请精简
高级检索

A novel enhancing MLP-SSO approach for predicting COVID-19 mortality time series

  • 期刊论文

  • 发表日期：

    2026年5月8日

查看全文
  • Shahryar Ghorbani,
  • Predrag S. Stanimirović,
  • Reza Rostamzadeh,
  • Hero Isavi,
  • Roshni Thendiyath

摘要

The COVID-19 pandemic has significantly altered global health, economies, and societal structures, making accurate daily case predictions essential for strategic planning and outbreak control. This study employs two machine learning models: a multi-layer perceptron (MLP) neural network and a bio-inspired sperm swarm optimization (SSO) algorithm hybridized with MLP (MLP-SSO), comparing their performance in forecasting daily COVID-19 deaths in Brazil, India, Russia, and the United States using time series data from January 20 to September 15, 2020. To ensure a realistic evaluation of the model’s forecasting capability, the data were chronologically split into training (first 70%) and test (remaining 30%) sets, preserving the temporal order of the time series. Additionally, the average mutual information method was utilized to determine optimal time lags, enabling the model to capture nonlinear relationships in the time series data. Model performance was evaluated using standard statistical metrics, including root mean square error (RMSE) and Pearson correlation coefficient (R). The results demonstrated that the MLP-SSO hybrid model outperformed the standalone MLP neural network across all four countries, significantly improving MLP performance. Specifically, the R increased from 0.8 to 0.94 in Brazil, from 0.78 to 0.92 in India, from 0.92 to 0.96 in Russia, and from 0.49 to 0.90 in the United States. Furthermore, the RMSE was reduced by 52.5% in Brazil, 9.3% in India, 84.0% in Russia, and 80.9% in the United States. These findings provide valuable insights for policymakers and health authorities in managing and controlling the COVID-19 pandemic.