Multi-Key Privacy-Preserving Training and Classification using Supervised Machine Learning Techniques in Cloud Computing

Abstract

Cloud computing contains lots of processing power and storage. Cloud computing and machine learning (ML) techniques enable large-scale data processing. The enhanced ML-based categorization technique is established in the cloud. However, there is a risk of privacy leaking of training data in the data processing. The computational and communication costs of the information possessor(s) must be maintained to a minimum. This study suggests a multi-key enhanced support vector machine (MK-FHE) and multi-key fully homomorphic encryption (MK-FHE) supervised machine learning method for encrypted data (ESVM).The results suggest that MK-FHE protects data privacy and is more effective in processing.