Publications (22)
Successive Refinement of Privacy
Antonious M. Girgis, Deepesh Data, Kamalika Chaudhuri +2
This work examines a novel question: how much randomness is needed to achieve local differential privacy (LDP)? A motivating scenario is providing {\em multiple levels of privacy}…
QuPeD: Quantized Personalization via Distillation with Applications to Federated Learning
Kaan Ozkara, Navjot Singh, Deepesh Data +1
Traditionally, federated learning (FL) aims to train a single global model while collaboratively using multiple clients and a server. Two natural challenges that FL algorithms face…
Data Encoding for Byzantine-Resilient Distributed Optimization
Deepesh Data, Linqi Song, Suhas Diggavi
We study distributed optimization in the presence of Byzantine adversaries, where both data and computation are distributed among worker machines, of which may be corrupt.…
SQuARM-SGD: Communication-Efficient Momentum SGD for Decentralized Optimization
Navjot Singh, Deepesh Data, Jemin George +1
In this paper, we propose and analyze SQuARM-SGD, a communication-efficient algorithm for decentralized training of large-scale machine learning models over a network. In SQuARM-SG…
How to Securely Compute the Modulo-Two Sum of Binary Sources
Deepesh Data, Bikash Kumar Dey, Manoj Mishra +1
In secure multiparty computation, mutually distrusting users in a network want to collaborate to compute functions of data which is distributed among the users. The users should no…
Interactive Secure Function Computation
Deepesh Data, Gowtham R. Kurri, Jithin Ravi +1
We consider interactive computation of randomized functions between two users with the following privacy requirement: the interaction should not reveal to either user any extra inf…
Communication and Randomness Lower Bounds for Secure Computation
Deepesh Data, Vinod M. Prabhakaran, Manoj M. Prabhakaran
In secure multiparty computation (MPC), mutually distrusting users collaborate to compute a function of their private data without revealing any additional information about their…
A Field Guide to Federated Optimization
Jianyu Wang, Zachary Charles, Zheng Xu +50
Federated learning and analytics are a distributed approach for collaboratively learning models (or statistics) from decentralized data, motivated by and designed for privacy prote…
Must the Communication Graph of MPC Protocols be an Expander?
Elette Boyle, Ran Cohen, Deepesh Data +1
Secure multiparty computation (MPC) on incomplete communication networks has been studied within two primary models: (1) Where a partial network is fixed a priori, and thus corrupt…
QuPeL: Quantized Personalization with Applications to Federated Learning
Kaan Ozkara, Navjot Singh, Deepesh Data +1
Traditionally, federated learning (FL) aims to train a single global model while collaboratively using multiple clients and a server. Two natural challenges that FL algorithms face…
Qsparse-local-SGD: Distributed SGD with Quantization, Sparsification, and Local Computations
Debraj Basu, Deepesh Data, Can Karakus +1
Communication bottleneck has been identified as a significant issue in distributed optimization of large-scale learning models. Recently, several approaches to mitigate this proble…
Flexible Accuracy for Differential Privacy
Aman Bansal, Rahul Chunduru, Deepesh Data +1
Differential Privacy (DP) has become a gold standard in privacy-preserving data analysis. While it provides one of the most rigorous notions of privacy, there are many settings whe…
Shuffled Model of Federated Learning: Privacy, Communication and Accuracy Trade-offs
Antonious M. Girgis, Deepesh Data, Suhas Diggavi +2
We consider a distributed empirical risk minimization (ERM) optimization problem with communication efficiency and privacy requirements, motivated by the federated learning (FL) fr…
SPARQ-SGD: Event-Triggered and Compressed Communication in Decentralized Stochastic Optimization
Navjot Singh, Deepesh Data, Jemin George +1
In this paper, we propose and analyze SPARQ-SGD, which is an event-triggered and compressed algorithm for decentralized training of large-scale machine learning models. Each node c…
On the Renyi Differential Privacy of the Shuffle Model
Antonious M. Girgis, Deepesh Data, Suhas Diggavi +2
The central question studied in this paper is Renyi Differential Privacy (RDP) guarantees for general discrete local mechanisms in the shuffle privacy model. In the shuffle model,…
On the Communication Complexity of Secure Computation
Deepesh Data, Vinod M. Prabhakaran, Manoj M. Prabhakaran
Information theoretically secure multi-party computation (MPC) is a central primitive of modern cryptography. However, relatively little is known about the communication complexity…
Renyi Differential Privacy of the Subsampled Shuffle Model in Distributed Learning
Antonious M. Girgis, Deepesh Data, Suhas Diggavi
We study privacy in a distributed learning framework, where clients collaboratively build a learning model iteratively through interactions with a server from whom we need privacy.…
Secure Computation of Randomized Functions: Further Results
Deepesh Data, Vinod M. Prabhakaran
We consider secure computation of randomized functions between two users, where both the users (Alice and Bob) have inputs, Alice sends a message to Bob over a rate-limited, noise-…
Byzantine-Resilient High-Dimensional Federated Learning
Deepesh Data, Suhas Diggavi
We study stochastic gradient descent (SGD) with local iterations in the presence of malicious/Byzantine clients, motivated by the federated learning. The clients, instead of commun…
Byzantine-Resilient SGD in High Dimensions on Heterogeneous Data
Deepesh Data, Suhas Diggavi
We study distributed stochastic gradient descent (SGD) in the master-worker architecture under Byzantine attacks. We consider the heterogeneous data model, where different workers…
A Generative Framework for Personalized Learning and Estimation: Theory, Algorithms, and Privacy
Kaan Ozkara, Antonious M. Girgis, Deepesh Data +1
A distinguishing characteristic of federated learning is that the (local) client data could have statistical heterogeneity. This heterogeneity has motivated the design of personali…
Secure Computation of Randomized Functions
Deepesh Data
Two user secure computation of randomized functions is considered, where only one user computes the output. Both the users are semi-honest; and computation is such that no user lea…