A design of predictive computational network for the analysis of fractional epidemical predictor-prey model. (December 2022)

• Record Type: Journal Article
• Authors: Shoaib, Muhammad; Abbasi, Aqsa Zafar; Raja, Muhammad Asif Zahoor; Nisar, Kottakkaran Sooppy
• Is Part Of: Chaos, solitons and fractals. Volume 165:Part 1(2022)
• Publication Date: 2022-12
• Subjects: Supervised machine learning -- Bayesian Regularization methodology -- Fractional epidemical model; Chaotic behavior in systems -- Periodicals; Solitons -- Periodicals; Fractals -- Periodicals; Chaotic behavior in systems; Fractals; Solitons; Periodicals; 003.7
• DOI: 10.1016/j.chaos.2022.112812
• Languages: English
• ISSNs: 0960-0779

Abstract

Abstract: Studying the dynamical system that represents the transmission of an outbreak from prey to predatory is crucial and significant because the results may be used to highlight a variety of real-world issues. Infections and predator-prey interplay have been linked to provide a complicated cumulative influence as regulators of prey and predator size range in predator-prey ecology. This paper presents a novel implementation of intelligent computation to analyze the dynamics of a fractional epidemiological model with disease infection in both the population (FEM-DIBP) using supervised machine learning (SMLs) optimized with the Bayesian Regularization methodology (BRM). The FOLotkaVoltera solver based on Grunwald-Letnikov is used to build the dataset for the FEM-DIBP for observation approach of the SML model of the system. Also state variables trajectories and Lorenz curves are constructed using FOLorenz to analyze the dynamics of the epidemical model. The BRM learned SMLs models are used in the train, test, and validating procedures to find the FEM-DIBP solutions to various situations depending on changing epidemical factors. The effectiveness of SML in solving FEMDIBP is demonstrated by estimated regression measurements, error histogram visualizations, and mean-squared error analyses. The suitability of the created BRMNNs for such a simulated situation is demonstrated by the most appealing mathematical findings for A.E within the range of 10 − 2 to 10 − 6 . Highlights: A novel implementation of intelligent computation to analyze the dynamics of a fractional epidemiological model Used supervised machine learning (SMLs) optimized with the Bayesian Regularization methodology The FOLotkaVoltera solver based on Grunwald-Letnikov is used to build the dataset. State variables trajectories and Lorenz curves are constructed to analyze the dynamics of the epidemical model.