Internal Audit Effectiveness for Data Compliance Using Privacy Federated Learning Model in Hospital Environment

Corporate governance privacy policies that follow the digital sovereignty, have an impact on firm’s value and it will reflect on M & A. The best practices and access to new technologies of data privacy policies after acquisition will be applied as per GDPR recommendations. The objective of this paper is to apply a privacy model to protect patients as per privacy policies and maintain cyber security data compliance. The checklist of the acquirer cyber security data compliance policy is applied to the acquired hospital by considering the GDPR articles. Federated learning privacy model is applied to protect and preserve the data during training and testing on OCT image dataset. The dataset was split for training and testing, and gaussian noise was added for privacy preservation from inversion attack. Further, to mitigate the label flipping attack, non-linear aggression method is applied and prediction was measured. Internal audit effectiveness for data compliance and noncompliance was explored and multiple linear regression model was applied to check the variance score. The variance score of 0.59 was obtained using the stimulated dataset and further it can be improved by considering the real time dataset.