ABSTRACT
Big data describes extraordinarily large datasets with complex and varied structures. These characteristics make it more difficult to store data, analyze them, use new techniques, or extract findings.
Big data analytics refers to the analysis of vast amounts of complex data to find hidden patterns or unrecognized correlations. However, the increasing use of big data and its privacy and security are in stark contrast. This chapter distinguishes between cybersecurity and privacy and focuses on issues with large-data privacy and security.
When big data grow, there is also a greater potential for someone’s privacy to be violated. Distributed systems are used because massive amounts of data require considerable processing power and storage. These systems enhance the potential for privacy breaches because numerous parties are involved. Data privacy issues are growing as machine learning is increasingly used in the Internet of Medical Things (IoMT) context. We offer an effective outsourced support vector machine (EPoSVM) approach that protects privacy and is intended for IoMT deployment.
We developed eight secure computation protocols to efficiently conduct basic floating-point and integer computations on a cloud server for secure support vector machine (SVM) training.
The proposed strategy guarantees the safety of the trained SVM model and protects the confidentiality of the training data. Security research has shown that EPoSVM and our recommended protocols adhere to security and privacy standards.
