Secure Multi-Party Computation

Bayatbabolghani, Fattaneh; Blanton, Marina

doi:10.1145/3243734.3264419

Cited by 21 publications

(11 citation statements)

References 25 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In this context, various algorithms have been used in FL systems, such as: Homomorphic encryption [ 39 ]: Users can compute and process the encrypted data without revealing the original data, and at the same time the user decrypted the processed data with the key, which is exactly the expected result. However, due to the additional encryption/decryption operations, homomorphic encryption incurs extremely high computational overhead; Secure multiparty computation (SMC) [ 62 ]: in this algorithm, the server is guaranteed to learn the parties’ inputs only in their entirety. However, SMC does not provide any confidentiality guarantee for the final model, which is still vulnerable to inference and model inversion attacks and can also be a reason for additional computational overhead.…”

Section: Federated Learningmentioning

confidence: 99%

“…Secure multiparty computation (SMC) [ 62 ]: in this algorithm, the server is guaranteed to learn the parties’ inputs only in their entirety. However, SMC does not provide any confidentiality guarantee for the final model, which is still vulnerable to inference and model inversion attacks and can also be a reason for additional computational overhead.…”

Section: Federated Learningmentioning

confidence: 99%

See 1 more Smart Citation

Reviewing Federated Machine Learning and Its Use in Diseases Prediction

Moshawrab

Adda

Bouzouane

et al. 2023

Sensors

View full text Add to dashboard Cite

Machine learning (ML) has succeeded in improving our daily routines by enabling automation and improved decision making in a variety of industries such as healthcare, finance, and transportation, resulting in increased efficiency and production. However, the development and widespread use of this technology has been significantly hampered by concerns about data privacy, confidentiality, and sensitivity, particularly in healthcare and finance. The “data hunger” of ML describes how additional data can increase performance and accuracy, which is why this question arises. Federated learning (FL) has emerged as a technology that helps solve the privacy problem by eliminating the need to send data to a primary server and collect it where it is processed and the model is trained. To maintain privacy and improve model performance, FL shares parameters rather than data during training, in contrast to the typical ML practice of sending user data during model development. Although FL is still in its infancy, there are already applications in various industries such as healthcare, finance, transportation, and others. In addition, 32% of companies have implemented or plan to implement federated learning in the next 12–24 months, according to the latest figures from KPMG, which forecasts an increase in investment in this area from USD 107 million in 2020 to USD 538 million in 2025. In this context, this article reviews federated learning, describes it technically, differentiates it from other technologies, and discusses current FL aggregation algorithms. It also discusses the use of FL in the diagnosis of cardiovascular disease, diabetes, and cancer. Finally, the problems hindering progress in this area and future strategies to overcome these limitations are discussed in detail.

show abstract

Section: Federated Learningmentioning

confidence: 99%

Section: Federated Learningmentioning

confidence: 99%

Reviewing Federated Machine Learning and Its Use in Diseases Prediction

Moshawrab

Adda

Bouzouane

et al. 2023

Sensors

View full text Add to dashboard Cite

show abstract

“…Secure multi-party computation (SMPC) [4] enables a set of parties 𝑃 = {𝑃 1 , ..., 𝑃 𝑚 }, where party 𝑃 𝑖 holds sensitive input 𝑥 𝑖 , to jointly compute a function 𝑦 = 𝑓 (𝑥 1 , ..., 𝑥 𝑚 ) while protecting each 𝑥 𝑖 . The computation needs to result in the correct value of 𝑦 (called the correctness property) and at the end of the computation each 𝑃 𝑖 learns nothing beyond 𝑦 (called the privacy property).…”

Section: Secure Multi-party Computationmentioning

confidence: 99%

PRICURE: Privacy-Preserving Collaborative Inference in a Multi-Party Setting

Jarin¹,

Eshete²

2021

Preprint

View full text Add to dashboard Cite

When multiple parties that deal with private data aim for a collaborative prediction task such as medical image classification, they are often constrained by data protection regulations and lack of trust among collaborating parties. If done in a privacy-preserving manner, predictive analytics can benefit from the collective prediction capability of multiple parties holding complementary datasets on the same machine learning task. This paper presents pricure, a system that combines complementary strengths of secure multi-party computation (SMPC) and differential privacy (DP) to enable privacypreserving collaborative prediction among multiple model owners. SMPC enables secret-sharing of private models and client inputs with non-colluding secure servers to compute predictions without leaking model parameters and inputs. DP masks true prediction results via noisy aggregation so as to deter a semi-honest client who may mount membership inference attacks. We evaluate pricure on neural networks across four datasets including benchmark medical image classification datasets. Our results suggest pricure guarantees privacy for tens of model owners and clients with acceptable accuracy loss. We also show that DP reduces membership inference attack exposure without hurting accuracy.

show abstract

“…A lightweight encryption scheme is one of the most important methods in privacy protection of IoT. It commonly uses encryption methods like homomorphic encryption [10] and secure multi-party computation (SMC) [11]. The invisibility of the original data and losslessness of the data are realized by the cryptography mechanism.…”

Section: A Data Privacy Protection In Iotmentioning

confidence: 99%

DynaPro: Dynamic Wireless Sensor Network Data Protection Algorithm in IoT via Differential Privacy

Liu

Huang

et al. 2019

IEEE Access

View full text Add to dashboard Cite

With the advent of the era of intelligent Internet of Things (IoT), more and more personal information are collected in the process of deploying IoT in various domains. This brings the problem of data protection in IoT. To address this problem, we propose a data protection algorithm, called DynaPro, based on differential privacy for dynamic wireless sensor networks (WSN), which is the sensor layer of IoT. DynaPro has addressed two issues in data protection of wireless sensor networks. The first issue is the dynamic topology of the network. In order to solve this problem, DynaPro has adopted methods like time window, hierarchical sampling, and graph similarity to process the snapshots of the dynamic network topology. The second issue is addition of noise in differential privacy. To address this issue, DynaPro uses Hierarchical Random Graph (HRG) as middleware to apply differential privacy protection. Instead of adding noise to network data and topology directly, DynaPro adds noise to HRG. In this way, DynaPro can hide the network topology but retain the necessary structural information to support data analysis. Theoretical analysis and experimental results show that DynaPro can preserve the important network features of the original network topology under the premise of the differential privacy protection model.

show abstract

Secure Multi-Party Computation

Cited by 21 publications

References 25 publications

Reviewing Federated Machine Learning and Its Use in Diseases Prediction

Reviewing Federated Machine Learning and Its Use in Diseases Prediction

PRICURE: Privacy-Preserving Collaborative Inference in a Multi-Party Setting

DynaPro: Dynamic Wireless Sensor Network Data Protection Algorithm in IoT via Differential Privacy

Contact Info

Product

Resources

About