Privacy-Preserving Homomorphic Encryption Schemes for Machine Learning in the Cloud (original) (raw)
This examination investigates the joining of protection safeguarding homomorphic encryption plans in cloud-based AI to address blossoming concerns regarding information security and protection. This proposed method is based on recent contributions and focuses on tailoring homomorphic encryption algorithms like Paillier and Fully Homomorphic Encryption (FHE) to specific machine learning tasks. To strike a balance between data utility and privacy, seamless compatibility with preprocessing pipelines is prioritized. Secure model preparation strategies, consolidating cryptographic conventions and secure conglomeration techniques, are crucial in saving the secrecy of delicate data. The improvement of encoded model assessment measurements guarantees a hearty evaluation of model execution without compromising protection. Our technique stretches out to strengthening the general security act through exhaustive examination and countermeasure execution, tending to likely weaknesses in homomorphic encryption. Coordination inside the cloud foundation is a focal subject, with an emphasis on versatility, similarity, and true relevance. Challenges connected with dormancy, asset utilization, and versatility to fluctuating jobs are addressed to exhibit the down-to-earth practicality of security safeguarding AI. Looking forward, future work ought to envelop headways in encryption calculations, client-driven contemplations, cooperation with industry partners, and novel applications in united learning and IoT situations.
Sign up for access to the world's latest research.
checkGet notified about relevant papers
checkSave papers to use in your research
checkJoin the discussion with peers
checkTrack your impact
Loading Preview
Sorry, preview is currently unavailable. You can download the paper by clicking the button above.