5 Jul 2019 · In this paper, we propose a method based on data encoding and error correction over real numbers to combat adversarial attacks.

Abstract: We consider distributed gradient computation, where both data and computation are distributed among m worker machines, t of which can be Byzantine ...

We propose Byzantine-resilient distributed optimization algorithms both for PGD and CD based on data encoding and error correction (over real numbers). As ...

to large-scale distributed learning [7]–[9]. In this paper, we propose a Byzantine-resilient distributed optimization algorithm based on data encoding and error.

One can also use computation redundancy [31]- [33] to achieve resilient convergence for distributed learning, which typically involves coding theory and ...

This work proposes a method based on data encoding and real error correction to combat the adversary, which can tolerate up to m/3 corrupt nodes and designs ...

, Data Encoding for Byzantine-Resilient Distributed Optimization. ; Data, Deepesh; Song, Linqi; Diggavi, Suhas N ;. , ISSN: 0018-9448 , 1557-9654; DOI ...

This paper does not make any assumptions on the probability distribution of agents' data points. We also note that all the prior works on Byzantine fault ...

2 Oct 2018 · Our method does not assume any probability distribution on the data. We develop a sparse encoding scheme which enables computationally efficient ...

<http://purl.org/dc/terms/title>. "data encoding methods for byzantine resilient distributed optimization"^^<http://www.w3.org/2001/XMLSchema#string>. 5. <http ...