Scholarlink

<p>This paper presents the numerical performance of the co-infection of dengue and coronavirus. The mathematical co-infection of dengue and coronavirus model is one of the nonlinear models. An implicit Runge-Kutta scheme is applied to find the dataset that reduces the mean square error by dividing the dataset into training data, 77%; testing data, 12%; and validation data, 11%. In this research, a soft computing Bayesian regularization neural network approach is proposed by using the activation logistic sigmoid function, construction of a hidden layer, twenty-two neurons, and the Bayesian regularization optimization approach, ensuring the procedure’s correctness by overlapping of outputs and reducing absolute error performances. Furthermore, some tests were performed in this research considering error histogram, regression coefficient, state transition values, and optimal training to develop the consistency of the stochastic machine learning neural network approach.</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 soft computing bayesian regularization neural network approach for the coendemic model of dengue disease and the coronavirus model

  • 期刊论文

  • 发表日期：

    2026年5月30日

查看全文
  • S. Sáenz-Sánchez,
  • J. F. Gómez-Aguilar,
  • R. F. Escobar-Jiménez,
  • Eduardo Pérez Careta,
  • J. E. Lavín-Delgado

摘要

This paper presents the numerical performance of the co-infection of dengue and coronavirus. The mathematical co-infection of dengue and coronavirus model is one of the nonlinear models. An implicit Runge-Kutta scheme is applied to find the dataset that reduces the mean square error by dividing the dataset into training data, 77%; testing data, 12%; and validation data, 11%. In this research, a soft computing Bayesian regularization neural network approach is proposed by using the activation logistic sigmoid function, construction of a hidden layer, twenty-two neurons, and the Bayesian regularization optimization approach, ensuring the procedure’s correctness by overlapping of outputs and reducing absolute error performances. Furthermore, some tests were performed in this research considering error histogram, regression coefficient, state transition values, and optimal training to develop the consistency of the stochastic machine learning neural network approach.