Securing Healthcare Data: Federated Learning for Privacy-Preserving AI in Medical Applications (Published)
Federated Learning (FL) is a technique used when sharing raw data cannot be done because of privacy laws. FL is used to train machine learning algorithms on decentralized data. Electronic health records, which hold private patient data, are one type of such data. In FL, local models are trained, and the model parameters are then combined on a central server instead of sharing sensitive data. But this approach poses privacy risks, so before disclosing the model parameters, privacy protection measures such data confidentiality must be put in existence. During the pandemic, there is a need to improve the healthcare system. Numerous advancements in Artificial Intelligence (AI) technology are continuously being utilized in several healthcare domains. Federated Learning (FL), one such development, has gained popularity mostly because of its decentralized, cooperative approach to creating AI models. Since integrating privacy algorithms can affect the utility, it is important to strike a balance when it comes to privacy and utility in FL research. The goal is to use strategies such as data generalizing, feature selection for reducing dimensions, and reduction in the confidentiality process to maximize FL’s effectiveness while preserving privacy. To create a predictive model for healthcare applications, this study also explores the idea of segmenting data based on attributes rather than records. It assesses the effectiveness of the model recommended by utilizing actual medical data.
Keywords: Data protection, Federated Learning, healthcare data security, medical applications, privacy-preserving AI, secure machine learning
Secure and Efficient Federated Learning Framework for Advanced Credit Card Fraud Detection with Optimization (Published)
In recent years, credit card fraud has cost banks and customers a great deal of money. A strong Fraud Detection System (FDS) is therefore necessary to reduce losses for consumers and banks. Our analysis shows that the dataset of credit card transactions is extremely biased, with many fewer examples of fraudulent purchases than of genuine ones. In addition, banks are typically prohibited from sharing their transaction statistics because of concerns about data security and privacy. These issues make it challenging for FDS to identify fraud tendencies and to identify them. In this investigation, we offer a framework in which we label FFD (Federated learning for Fraud Detection) to train a fraud detection model utilizing behavior features with federated learning and convolutional neural networks (CNN) with Greylag Goose Optimization. In contrast to the conventional FDS trained on cloud-centralized data, FFD allows banks to use training data from their local databases to create fraud detection models. Subsequently, a shared fraud detection model is created by combining locally computed updates. Banks can profit collectively from a shared model without exchanging datasets to safeguard the cardholders’ sensitive information. In addition, an oversampling strategy is employed to counterbalance the skewed dataset. We use an extensive set of actual credit card transactions to assess the effectiveness of our credit card FDS system. The findings demonstrate the great accuracy with which each algorithm may be applied to the detection of credit card fraud.
Keywords: CNN, Credit Card Fraud Detection, Efficiency, Federated Learning, Graylag Goose Optimization, Security