On Lightweight Privacy-preserving Collaborative Learning for Internet of Things by Independent Random Projections

Abstract

The Internet of Things (IoT) will be a main data generation infrastructure for achieving better system intelligence. This article considers the design and implementation of a practical privacy-preserving collaborative learning scheme, in which a curious learning coordinator trains a better machine learning model based on the data samples contributed by a number of IoT objects, while the confidentiality of the raw forms of the training data is protected against the coordinator. Existing distributed machine learning and data encryption approaches incur significant computation and communication overhead, rendering them ill-suited for resource-constrained IoT objects. We study an approach that applies independent random projection at each IoT object to obfuscate data and trains a deep neural network at the coordinator based on the projected data from the IoT objects. This approach introduces light computation overhead to the IoT objects and moves most workload to the coordinator that can have sufficient computing resources. Although the independent projections performed by the IoT objects address the potential collusion between the curious coordinator and some compromised IoT objects, they significantly increase the complexity of the projected data. In this article, we leverage the superior learning capability of deep learning in capturing sophisticated patterns to maintain good learning performance. Extensive comparative evaluation shows that this approach outperforms other lightweight approaches that apply additive noisification for differential privacy and/or support vector machines for learning in the applications with light to moderate data pattern complexities.

Application of Internet of Things Audio Technology Based on Parallel Storage System in Music Classroom

“…It can be obtained from formula (10) that C is a (Hermitian) Toeplitz matrix. We determine by M complex numbers that for some u ∈ ℂ M can be written as C = TðuÞ, where (11), a characterization such as formula ( 12) can be derived.…”

“…Students' concentration can be improved through interaction, and students can also directly experience the delicate changes in music through personal demonstrations, but they are occasionally limited by the epidemic. In short, online education is convenient and fragmented, allowing learners to join learning anytime, anywhere; traditional education systems are efficient and conducive to learners' scientific and systematic training [10]. In addition, the concept of managers must be updated.…”

A Survey of Machine and Deep Learning Methods for Privacy Protection in the Internet of Things

“…Accuracy is slightly reduced compared to the centralized framework but communication and computation costs are significantly reduced. Other solutions consider Gaussian projections [ 90 ] to improve performance of privacy-preserving collaborative learning. IoT resource-constrained participants apply independent multiplicative Gaussian Random Projections (GRP) on training data vectors to obfuscate the contributed training data.…”

Integrating IoT-Sensing and Crowdsensing with Privacy: Privacy-Preserving Hybrid Sensing for Smart Cities

“…In the literature, the typical solutions are data perturbation and cryptography. For example, [22] implemented a privacy-preserving collaborative learning model, using a multiplicative random projection approach to obfuscate training data at each resource-limited IoT object. The authors claimed that their approach outperformed the other approaches based on additive noisification.…”