SCaO-DMN: Skill Cavier Optimization Enabled Deep Maxout Network for Privacy Preservation of Health Care Data based Federated Learning

Privacy for cloud properties is significant since the cloud has entrusted, and isolated sensitive data for public access. Thus, the Skill Cavier Optimization-based Deep Maxout Network (SCaO-DMN) is developed for privacy preservation. Here, the local training is done with the local data at each node and is updated on the server. Then, model aggregation is conducted and downloaded by a global module at each node. In the training module, the healthcare data is subjected to noisy treatment for privacy preservation. After that, the noise matrix process is achieved by DMN trained by SaCO. Lastly, Local updation and aggregation at the server is done using the average and stored in the cloud. The devised model is evaluated using Accuracy, loss function and Mean Squared Error (MSE) and obtained values of 91.887, 0.081, and 0.012. The SCaO-DMN can help in an effective treatment plan as it protects the data more securely.