Byzantine-resilient non-convex stochastic gradient descent
We study adversary-resilient stochastic distributed optimization, in which $ m $ machines
can independently compute stochastic gradients, and cooperate to jointly optimize over their …
can independently compute stochastic gradients, and cooperate to jointly optimize over their …
Robust training in high dimensions via block coordinate geometric median descent
Geometric median (GM) is a classical method in statistics for achieving robust estimation of
the uncorrupted data; under gross corruption, it achieves the optimal breakdown point of 1/2 …
the uncorrupted data; under gross corruption, it achieves the optimal breakdown point of 1/2 …
Byzantine-robust and communication-efficient distributed non-convex learning over non-IID data
Motivated by the emerging federated learning applications, we jointly consider the problems
of Byzantine-robustness and communication efficiency in distributed non-convex learning …
of Byzantine-robustness and communication efficiency in distributed non-convex learning …
C-RSA: Byzantine-robust and communication-efficient distributed learning in the non-convex and non-IID regime
The emerging federated learning applications raise challenges of Byzantine-robustness and
communication efficiency in distributed non-convex learning over non-IID data. To address …
communication efficiency in distributed non-convex learning over non-IID data. To address …
Byzantine resilient non-convex SCSG with distributed batch gradient computations
Distributed learning is an important paradigm in the current machine learning algorithms
with large datasets. In this paper, distributed stochastic optimization problem of minimizing a …
with large datasets. In this paper, distributed stochastic optimization problem of minimizing a …
Asynchronous SGD with stale gradient dynamic adjustment for deep learning training
Asynchronous stochastic gradient descent (ASGD) is a computationally efficient algorithm,
which speeds up deep learning training and plays an important role in distributed deep …
which speeds up deep learning training and plays an important role in distributed deep …
Communication-efficient distributed eigenspace estimation with arbitrary node failures
V Charisopoulos, A Damle - Advances in Neural …, 2022 - proceedings.neurips.cc
We develop an eigenspace estimation algorithm for distributed environments with arbitrary
node failures, where a subset of computing nodes can return structurally valid but otherwise …
node failures, where a subset of computing nodes can return structurally valid but otherwise …
On Robust Machine Learning in the Presence of Adversaries
S Bulusu - 2023 - search.proquest.com
In today's highly connected world, the number of smart devices worldwide has increased
exponentially. These devices generate huge amounts of real-time data, perform complicated …
exponentially. These devices generate huge amounts of real-time data, perform complicated …
Arbitrarily Accurate Aggregation Scheme for Byzantine SGD
A Maurer - 25th International Conference on Principles of …, 2022 - drops.dagstuhl.de
A very common optimization technique in Machine Learning is Stochastic Gradient Descent
(SGD). SGD can easily be distributed: several workers try to estimate the gradient of a loss …
(SGD). SGD can easily be distributed: several workers try to estimate the gradient of a loss …
COMPUTATIONALLY EFFICIENT AND ROBUST METHODS FOR LARGE-SCALE OPTIMIZATION AND SCIENTIFIC COMPUTING
V Charisopoulos - 2023 - ecommons.cornell.edu
This thesis is concerned with the design and analysis of computationally efficient algorithms
for large-scale optimization and scientific computing. It aims to address two primary …
for large-scale optimization and scientific computing. It aims to address two primary …