ABSTRACT
In recent years, the Internet of Things (IoT) has become a dominant data generation framework for establishing a higher level of system intelligence. At the same time, to avail the full advantage of this domain, the adopters of IoT are also keen on applying Machine Learning (ML) algorithms to these datasets to reveal new business insights. However, these datasets contain sensitive information that demands careful processing to prevent privacy breaches. Many existing privacy-preserving ML (PPML) algorithms are unsuitable for these resource-constrained devices. We propose a novel PPML technique that can be executed on IoT devices using the Bloom Filter encoded IoT dataset in XNOR-Net architecture. The preliminary experimental result using the MNIST dataset shows satisfactory performance.
- Burton H Bloom. 1970. Space/time trade-offs in hash coding with allowable errors. Commun. ACM 13, 7 (1970), 422--426.Google ScholarDigital Library
- David Cash, Paul Grubbs, Jason Perry, and Thomas Ristenpart. 2015. Leakageabuse attacks against searchable encryption. In Proceedings of the 22nd ACM SIGSAC conference on computer and communications security. 668--679.Google ScholarDigital Library
- Li Deng. 2012. The mnist database of handwritten digit images for machine learning research [best of the web]. IEEE Signal Processing Magazine 29, 6 (2012), 141--142.Google ScholarCross Ref
- Úlfar Erlingsson, Vasyl Pihur, and Aleksandra Korolova. 2014. Rappor: Randomized aggregatable privacy-preserving ordinal response. In Proceedings of the 2014 ACM SIGSAC conference on computer and communications security. 1054--1067.Google ScholarDigital Library
- Craig Gentry. 2009. A fully homomorphic encryption scheme. Stanford university.Google Scholar
- Itay Hubara, Matthieu Courbariaux, Daniel Soudry, Ran El-Yaniv, and Yoshua Bengio. 2016. Binarized neural networks. Advances in neural information processing systems 29 (2016).Google Scholar
- Yann LeCun et al. 2015. LeNet-5, convolutional neural networks. URL: http://yann.lecun. com/exdb/lenet 20, 5 (2015), 14.Google Scholar
- Mohammad Rastegari, Vicente Ordonez, Joseph Redmon, and Ali Farhadi. 2016. Xnor-net: Imagenet classification using binary convolutional neural networks. In European conference on computer vision. Springer, 525--542.Google ScholarCross Ref
- Mohammad Rastegari, Vicente Ordonez, Joseph Redmon, and Ali Farhadi. 2020. Enabling AI at the edge with XNOR-networks. Commun. ACM 63, 12 (2020), 83--90.Google ScholarDigital Library
- Taylor Simons and Dah-Jye Lee. 2019. A review of binarized neural networks. Electronics 8, 6 (2019), 661.Google ScholarCross Ref
- Dinusha Vatsalan and Peter Christen. 2016. Privacy-preserving matching of similar patients. Journal of biomedical informatics 59 (2016), 285--298.Google ScholarDigital Library
- Wanli Xue, Dinusha Vatsalan, Wen Hu, and Aruna Seneviratne. 2020. Sequence data matching and beyond: New privacy-preserving primitives based on Bloom filters. IEEE Transactions on Information Forensics and Security 15 (2020), 2973--2987.Google ScholarCross Ref
- Mengyao Zheng, Dixing Xu, Linshan Jiang, Chaojie Gu, Rui Tan, and Peng Cheng. 2019. Challenges of privacy-preserving machine learning in IoT. In Proceedings of the First International Workshop on Challenges in Artificial Intelligence and Machine Learning for Internet of Things. 1--7.Google ScholarDigital Library
- Liehuang Zhu, Xiangyun Tang, Meng Shen, Feng Gao, Jie Zhang, and Xiaojiang Du. 2021. Privacy-preserving machine learning training in IoT aggregation scenarios. IEEE Internet of Things Journal 8, 15 (2021), 12106--12118.Google ScholarCross Ref
Index Terms
- BloomXNOR-Net: privacy-preserving machine learning in IoT
Recommendations
A decision-support framework for data anonymization with application to machine learning processes
Highlights- A methodology to identify data anonymization strategies providing suitable trade-offs between privacy and data utilization.
AbstractThe application of machine learning techniques to large and distributed data archives might result in the disclosure of sensitive information about the data subjects. Data often contain sensitive identifiable information, and even if ...
Taking back control of privacy: a novel framework for preserving cloud-based firewall policy confidentiality
As the cloud computing paradigm evolves, new types of cloud-based services have become available, including security services. Some of the most important and most commonly adopted security services are firewall services. These cannot be easily deployed ...
Privacy Preserving Face Recognition Utilizing Differential Privacy
AbstractFacial recognition technologies are implemented in many areas, including but not limited to, citizen surveillance, crime control, activity monitoring, and facial expression evaluation. However, processing biometric information is a ...
Comments