Reaching Data Confidentiality and Model Accountability on the CalTrain

Gu, Zhongshu; Jamjoom, Hani; Su, Dong; Huang, Heqing; Zhang, Jialong; Ma, Tengfei; Pendarakis, Dimitrios; Molloy, Ian

doi:10.1109/dsn.2019.00044

Cited by 21 publications

(13 citation statements)

References 36 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This means that the approaches assume that labels exist for all of the data in the federated network. However, in practice, the data generated in the network may be unlabeled or mislabeled [197]. This poses a big challenge to the server to find participants with appropriate data for model training.…”

Section: Challenges and Future Research Directionsmentioning

confidence: 99%

Federated Learning in Mobile Edge Networks: A Comprehensive Survey

Lim

Luong

Hoang

et al. 2020

IEEE Commun. Surv. Tutorials

1,630

797

View full text Add to dashboard Cite

In recent years, mobile devices are equipped with increasingly advanced sensing and computing capabilities. Coupled with advancements in Deep Learning (DL), this opens up countless possibilities for meaningful applications, e.g., for medical purposes and in vehicular networks. Traditional cloudbased Machine Learning (ML) approaches require the data to be centralized in a cloud server or data center. However, this results in critical issues related to unacceptable latency and communication inefficiency. To this end, Mobile Edge Computing (MEC) has been proposed to bring intelligence closer to the edge, where data is produced. However, conventional enabling technologies for ML at mobile edge networks still require personal data to be shared with external parties, e.g., edge servers. Recently, in light of increasingly stringent data privacy legislations and growing privacy concerns, the concept of Federated Learning (FL) has been introduced. In FL, end devices use their local data to train an ML model required by the server. The end devices then send the model updates rather than raw data to the server for aggregation. FL can serve as an enabling technology in mobile edge networks since it enables the collaborative training of an ML model and also enables DL for mobile edge network optimization. However, in a large-scale and complex mobile edge network, heterogeneous devices with varying constraints are involved. This raises challenges of communication costs, resource allocation, and privacy and security in the implementation of FL at scale. In this survey, we begin with an introduction to the background and fundamentals of FL. Then, we highlight the aforementioned challenges of FL implementation and review existing solutions. Furthermore, we present the applications of FL for mobile edge network optimization. Finally, we discuss the important challenges and future research directions in FL.

show abstract

Section: Challenges and Future Research Directionsmentioning

confidence: 99%

Federated Learning in Mobile Edge Networks: A Comprehensive Survey

Lim

Luong

Hoang

et al. 2020

IEEE Commun. Surv. Tutorials

1,630

797

View full text Add to dashboard Cite

show abstract

“…The process uses predefined approaches and stores or distributes information to the intended users for analysis. Although collaborative data training breaks data monopoly, it also raises new challenges of model accountability, privacy protection, and data confidentiality (Gu et al, 2019). Although various mechanisms have been implemented to enhance data privacy and confidentiality, data breach incidents expose students' sensitive data (Hlioui et al, 2021).…”

Section: Confidentiality and Privacy Issuesmentioning

confidence: 99%

“…Although various mechanisms have been implemented to enhance data privacy and confidentiality, data breach incidents expose students' sensitive data (Hlioui et al, 2021). Besides, most government regulations and laws on personal privacy protection do not permit privacy-sensitive and mission-critical domains such as education institutions to share raw data with third parties (Gu et al, 2019). The policy poses a significant challenge in implementing PAAs.…”

Section: Confidentiality and Privacy Issuesmentioning

confidence: 99%

Mitigating Academic Institution Dropout Rates with Predictive Analytics Algorithms

Lainjo¹

2023

IJETS

View full text Add to dashboard Cite

Multiple studies show student dropout rates across rich, low and middle-income countries. With the development of machine learning, various learning institutions have initiated predictive analytics algorithms to solve this issue. However, there is limited knowledge on how this strategy has been employed to optimize retention rates. The research aims to comprehensively review PAAs Strategies and demonstrate how they have been effectively implemented in optimizing retention rates with a case study focusing on North America. The study employed a quantitative research methodology and thematic study design to review the existing literature. The result shows that learning institutions used a combination of data mining classifiers such as k-Nearest Neighbor, Neural Networks, Decision Tree, and Naive Bayes to categorize student dropout rates and optimize retention rates. The PAAs approach has been utilized by various educational institutions across North America, applying multiple predicting algorithms to ensure efficiency. Keywords: Predictive analytics algorithms, mitigating dropout rates, retention rates, classifiers, Learning institutions, North America

show abstract

“…TEEs can act as trustworthy intermediaries for isolating and orchestrating ML processes and replace the expensive cryptographic primitives. For example, Software Guard Extensions (SGX) has been leveraged to support secure model inference [15,52], privacy-preserving multiparty machine learning [16,20,21,40,43], and analytics on sensitive data [4,9,46,64]. However, TEEs are not panacea to address all trust problems.…”

Section: Trusted Execution Environmentsmentioning

confidence: 99%

Separation of Powers in Federated Learning

Cheng,

Eykholt,

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Federated Learning (FL) enables collaborative training among mutually distrusting parties. Model updates, rather than training data, are concentrated and fused in a central aggregation server. A key security challenge in FL is that an untrustworthy or compromised aggregation process might lead to unforeseeable information leakage. This challenge is especially acute due to recently demonstrated attacks that have reconstructed large fractions of training data from ostensibly "sanitized" model updates.In this paper, we introduce Truda, a new cross-silo FL system, employing a trustworthy and decentralized aggregation architecture to break down information concentration with regard to a single aggregator. Based on the unique computational properties of model-fusion algorithms, all exchanged model updates in Truda are disassembled at the parameter-granularity and re-stitched to random partitions designated for multiple TEE-protected aggregators. Thus, each aggregator only has a fragmentary and shuffled view of model updates and is oblivious to the model architecture. Our new security mechanisms can fundamentally mitigate training data reconstruction attacks, while still preserving the final accuracy of trained models and keeping performance overheads low.

show abstract

Reaching Data Confidentiality and Model Accountability on the CalTrain

Cited by 21 publications

References 36 publications

Federated Learning in Mobile Edge Networks: A Comprehensive Survey

Federated Learning in Mobile Edge Networks: A Comprehensive Survey

Mitigating Academic Institution Dropout Rates with Predictive Analytics Algorithms

Separation of Powers in Federated Learning

Contact Info

Product

Resources

About