Abstract
In this paper, a fractional order model based on the management of waste plastic in the ocean (FO-MWPO) is numerically investigated. The mathematical form of the FO-MWPO model is categorized into three components, waste plastic, Marine debris, and recycling. The stochastic numerical solvers using the Levenberg-Marquardt backpropagation neural networks (LMQBP-NNs) have been applied to present the numerical solutions of the FO-MWPO system. The competency of the method is tested by taking three variants of the FO-MWPO model based on the fractional order derivatives. The data ratio is provided for training, testing and authorization is 77%, 12%, and 11% respectively. The exactness of LMQBP-NNs is observed by using the comparative performances of the obtained and the Adams-BashforthMoulton method. To verify the competence, validity, capability, exactness, and consistency of LMQBP-NNs, the performances have been obtained using the regression, state transitions, error histograms, correlation and mean square error.