Deep neural networks simulation for analysing rheological characteristics of Ag+Au/Blood Casson-Williamson hybrid nanofluid flow over a rotating disk: An application to targeted drug delivery

➢ The heat transfer analysis of a MHD Casson-Williamson hybrid nanofluid flow across a spinning porous disk is quantitatively investigated. ➢ The effects of a non-linear thermal radiation, viscosity and porous medium, and Casson-Williamson fluid are investigated in the present study. ➢ The well-organized Levenberg-Marquardt algorithm with Artificial neural networks (LMA-DNNs) is presented to solve differential systems ascending in the planned nanofluid and hybrid nanofluid cases. ➢ The computational solutions of mathematical problems deriving from the Casson- Williamson Au-Ag/Blood model are made more accurate with its improved framework. ➢ The suggested technique has decisive characteristics, including reliability and permanence as well as uniformity with flawless execution. ➢ A number of statistically important metrics, such as histogram analysis and MSE analysis are used to validate the claimed precision of the algorithm that was designed. Deep neural networks (DNNs) remarkable ability to handle highly complex and difficult mathematical problems is a major reason for their remarkable appeal. Computer deep neural network methodologies are applied in this research for the purpose of analysing the transfer analysis of magnetohydrodynamics Casson-Williamson hybrid nanofluid flow over a rotating disk along with a porous medium and nonlinear thermal radiation. The highly non-linear governing equations are converted PDEs into ODEs with the help of similarity variables. The present article considered base fluid as blood and the corresponding nanoparticles are Au and Ag. After that, by using the Differential Transform Method by generating graphs and tables. The comparison of total performance of nanofluids (R 2 = 0.97824) and hybrid nanofluids (R 2 = 0.98126), shows that the hybrid nanofluid case is superior to the nanofluid case. The accuracy of the semi-analytical solutions is further improved by combining them with a Deep Neural Network, which enhances their usefulness for various predictive modelling applications in biological sciences, biological computing, and fluid mechanics. The present study demonstrates the below key points: • Non-linear dynamics analysis on Casson-Williamson hybrid nanofluid flow over a rotating disk. • The present research analysed the effects of magnetohydrodynamic, porous medium, and non-linear thermal radiation. • The well-organized Deep neural networks (DNNs) are presented to solve differential systems ascending in the planned nanofluid and hybrid nanofluid cases.