Privacy-preserving collaborative machine learning on genomic data using TensorFlow

Cheng, Hong; Huang, Zhicong; Lu, Wenjie; Qu, Hunter; Ma, Li; Dahl, Morten; Mancuso, Jason

doi:10.1145/3393527.3393535

Cited by 10 publications

(6 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…An example of how SMPC protocols and SMPC-supporting machine learning libraries can be used is shown in the study by Hong et al [ 284 ], which used TF Encrypted to train a classifier on 2 genomic data sets, each containing a large number of features (12,634 and 17,814 features per sample), to detect tumors as part of the iDASH challenge. This task had an additional challenge because the 2 data sets were heavily imbalanced, but common countermeasures to this are difficult to implement in an SMPC framework.…”

Section: Resultsmentioning

confidence: 99%

Data Anonymization for Pervasive Health Care: Systematic Literature Mapping Study

Zuo¹,

Watson²,

Budgen³

et al. 2021

JMIR Med Inform

View full text Add to dashboard Cite

Background Data science offers an unparalleled opportunity to identify new insights into many aspects of human life with recent advances in health care. Using data science in digital health raises significant challenges regarding data privacy, transparency, and trustworthiness. Recent regulations enforce the need for a clear legal basis for collecting, processing, and sharing data, for example, the European Union’s General Data Protection Regulation (2016) and the United Kingdom’s Data Protection Act (2018). For health care providers, legal use of the electronic health record (EHR) is permitted only in clinical care cases. Any other use of the data requires thoughtful considerations of the legal context and direct patient consent. Identifiable personal and sensitive information must be sufficiently anonymized. Raw data are commonly anonymized to be used for research purposes, with risk assessment for reidentification and utility. Although health care organizations have internal policies defined for information governance, there is a significant lack of practical tools and intuitive guidance about the use of data for research and modeling. Off-the-shelf data anonymization tools are developed frequently, but privacy-related functionalities are often incomparable with regard to use in different problem domains. In addition, tools to support measuring the risk of the anonymized data with regard to reidentification against the usefulness of the data exist, but there are question marks over their efficacy. Objective In this systematic literature mapping study, we aim to alleviate the aforementioned issues by reviewing the landscape of data anonymization for digital health care. Methods We used Google Scholar, Web of Science, Elsevier Scopus, and PubMed to retrieve academic studies published in English up to June 2020. Noteworthy gray literature was also used to initialize the search. We focused on review questions covering 5 bottom-up aspects: basic anonymization operations, privacy models, reidentification risk and usability metrics, off-the-shelf anonymization tools, and the lawful basis for EHR data anonymization. Results We identified 239 eligible studies, of which 60 were chosen for general background information; 16 were selected for 7 basic anonymization operations; 104 covered 72 conventional and machine learning–based privacy models; four and 19 papers included seven and 15 metrics, respectively, for measuring the reidentification risk and degree of usability; and 36 explored 20 data anonymization software tools. In addition, we also evaluated the practical feasibility of performing anonymization on EHR data with reference to their usability in medical decision-making. Furthermore, we summarized the lawful basis for delivering guidance on practical EHR data anonymization. Conclusions This systematic literature mapping study indicates that anonymization of EHR data is theoretically achievable; yet, it requires more research efforts in practical implementations to balance privacy preservation and usability to ensure more reliable health care applications.

show abstract

Section: Resultsmentioning

confidence: 99%

Data Anonymization for Pervasive Health Care: Systematic Literature Mapping Study

Zuo¹,

Watson²,

Budgen³

et al. 2021

JMIR Med Inform

View full text Add to dashboard Cite

show abstract

“…[6,35,4,7,24] developed new protocols based on arithmetic and boolean sharing under 3-server or 4server settings and claimed to outperform previous methods. Open-sourced PPML libraries, such as CryptFlow [25], TF-Encrypted [20], and Crypten [23] are also based on the semihonest 3-servers setting where a certain party is usually used to generate Beaver triples.…”

Section: Cryptographic Methodsmentioning

confidence: 99%

“…Here we describe an example of shifting error. Let = 64, = 20, is a result of multiplication before shifting, and 20 . Clearly, the desired product is (2 63 + 2 20 ) × 2 mod 2 64 2 20 = 2 (since it is performed on integer ring ℤ 2 64 ).…”

Section: Fixed-point Arithmeticmentioning

confidence: 99%

See 1 more Smart Citation

Towards Secure and Practical Machine Learning via Secret Sharing and Random Permutation

Zheng¹,

Chen²,

Zheng³

et al. 2021

Preprint

View full text Add to dashboard Cite

With the increasing demands for privacy protection, privacy-preserving machine learning has been drawing much attention in both academia and industry. However, most existing methods have their limitations in practical applications. On the one hand, although most cryptographic methods are provable secure, they bring heavy computation and communication. On the other hand, the security of many relatively efficient private methods (e.g., federated learning and split learning) is being questioned, since they are non-provable secure. Inspired by previous work on privacy-preserving machine learning, we build a privacy-preserving machine learning framework by combining random permutation and arithmetic secret sharing via our compute-after-permutation technique. Since our method reduces the cost for element-wise function computation, it is more efficient than existing cryptographic methods. Moreover, by adopting distance correlation as a metric for privacy leakage, we demonstrate that our method is more secure than previous non-provable secure methods. Overall, our proposal achieves a good balance between security and efficiency. Experimental results show that our method not only is up to 6× faster and reduces up to 85% network traffic compared with state-ofthe-art cryptographic methods, but also leaks less privacy during the training process compared with non-provable secure methods.

show abstract

“…Although so many solutions have been proposed to solve the imbalanced data sets problem, the privacy-preserving issue has not been well resolved. To the best of our knowledge, Hong et al [25] proposed a secure collaborative machine-learning solution in which they used secure multiparty computation to adjust the class weight for the imbalanced dataset. at is, the privacy-preserving issue of the imbalanced dataset was tackled at the machine-learning algorithm level.…”

Section: Introductionmentioning

confidence: 99%

Privacy-Preserving Minority Oversampling Protocols with Fully Homomorphic Encryption

Sun

Yang

Liu

2022

Security and Communication Networks

View full text Add to dashboard Cite

In recent years, blockchain and machine-learning techniques have received increasing attention both in theoretical and practical aspects. However, the applications of these techniques have many challenges, one of which is the privacy-preserving issue. In this paper, we focus on, specifically, the privacy-preserving issue of imbalanced datasets, a commonly found problem in real-world applications. Built based on the fully homomorphic encryption technique, this paper presents two new secure protocols, Privacy-Preserving Synthetic Minority Oversampling Protocol (PPSMOS) and Borderline Privacy-Preserving Synthetic Minority Oversampling Protocol (Borderline-PPSMOS). Our analysis reveals that PPSMOS is generally more efficient in performance than Borderline-PPSMOS. However, Borderline-PPSMOS achieves a better TP rate and F-Value than PPSMOS.

show abstract

Privacy-preserving collaborative machine learning on genomic data using TensorFlow

Cited by 10 publications

References 12 publications

Data Anonymization for Pervasive Health Care: Systematic Literature Mapping Study

Data Anonymization for Pervasive Health Care: Systematic Literature Mapping Study

Towards Secure and Practical Machine Learning via Secret Sharing and Random Permutation

Privacy-Preserving Minority Oversampling Protocols with Fully Homomorphic Encryption

Contact Info

Product

Resources

About