In this investigation, we fill in missing values with the use of a unique sequential approach to data scaling that combines resilient scaling, z-standardization, and min-max scaling. Next, m-Xception does the classification by employing a different architecture than the initial convolutions. Furthermore, the convolution layer is broken down into depth-based sub-layers that are linked together by linear residuals. This productive model was trained using a two-stage transfer learning strategy. The kernel values of the proposed model are optimally selected for large-scale cases with the use of a Squeaky Wheel Optimisation (SWO) metaheuristic. The projected model was tested by simulation on the canonical CKD dataset and assessed statistically. The findings suggest that a fully automated method of assessing CKD severity is feasible. These results give support for the hypothesis that a unique approach to problem resolution may be achieved by combining predictive modelling with the most recent deep learning developments. This may be tested in the context of renal illness and beyond..

Squeaky Wheel Optimization; Chronic Kidney Disease; Cloud Computing; Min-Max Scaling. Modified-Xception;