English (United Kingdom)

https://curated-unify.zendy.io/wp-json/zendy-region/v1/featured_content/oa?rat=en

https://curated-unify.zendy.io/wp-json/zendy-region/v1/highlighted_journal/

Global: Zendy Plus annual

Presents the access of premium content as premium feature

Premium Content

Presents the keyphrase highlighting as premium feature

Keyphrase Highlighting

Presents the summarisation as premium feature

Summarisation

Insights

Presents the pdf analysis as premium feature

PDF Analysis

Presents the zaia usage as premium feature

ZAIA

Global: Zendy Plus monthly

Global: Zendy Tools annual

Global: Zendy Tools monthly

Global: Zendy Free Plan

With the popularity of cloud computing and machine learning, it has been atrend to outsource machine learning processes (including model training andmodel-based inference) to cloud. By the outsourcing, other than utilizing theextensive and scalable resource offered by the cloud service provider, it willalso be attractive to users if the cloud servers can manage the machinelearning processes autonomously on behalf of the users. Such a feature will beespecially salient when the machine learning is expected to be a long-termcontinuous process and the users are not always available to participate. Dueto security and privacy concerns, it is also desired that the autonomouslearning preserves the confidentiality of users' data and models involved.Hence, in this paper, we aim to design a scheme that enables autonomous andconfidential model refining in cloud. Homomorphic encryption and trustedexecution environment technology can protect confidentiality for autonomouscomputation, but each of them has their limitations respectively and they arecomplementary to each other. Therefore, we further propose to integrate thesetwo techniques in the design of the model refining scheme. Throughimplementation and experiments, we evaluate the feasibility of our proposedscheme. The results indicate that, with our proposed scheme the cloud servercan autonomously refine an encrypted model with newly provided encryptedtraining data to continuously improve its accuracy. Though the efficiency isstill significantly lower than the baseline scheme that refines plaintext-modelwith plaintext-data, we expect that it can be improved by fully utilizing thehigher level of parallelism and the computational power of GPU at the cloudserver.

Integrating Homomorphic Encryption and Trusted Execution Technology for Autonomous and Confidential Model Refining in Cloud