arXiv subjects

Albin Severinson

Publications and source records attributed to Albin Severinson.

5 recordsLinked to original sources

arXiv2021

DSAG: A mixed synchronous-asynchronous iterative method for straggler-resilient learning

We consider straggler-resilient learning. In many previous works, e.g., in the coded computing literature, straggling is modeled as random delays that are independent and identically distributed between workers. However, in many practical scenarios, a given worker may straggle over an extended period of time. We propose a latency model that captures this behavior and is substantiated by traces collected on Microsoft Azure, Amazon Web Services (AWS), and a small local cluster. Building on this model, we propose DSAG, a mixed synchronous-asynchronous iterative optimization method, based on the stochastic average gradient (SAG) method, that combines timely and stale results. We also propose a dynamic load-balancing strategy to further reduce the impact of straggling workers. We evaluate DSAG for principal component analysis, cast as a finite-sum optimization problem, of a large genomics dataset, and for logistic regression on a cluster composed of 100 workers on AWS, and find that DSAG is up to about 50% faster than SAG, and more than twice as fast as coded computing methods, for the particular scenario that we consider.

cs.DC↗

arXiv2019

Coded Distributed Tracking

We consider the problem of tracking the state of a process that evolves over time in a distributed setting, with multiple observers each observing parts of the state, which is a fundamental information processing problem with a wide range of applications. We propose a cloud-assisted scheme where the tracking is performed over the cloud. In particular, to provide timely and accurate updates, and alleviate the straggler problem of cloud computing, we propose a coded distributed computing approach where coded observations are distributed over multiple workers. The proposed scheme is based on a coded version of the Kalman filter that operates on data encoded with an erasure correcting code, such that the state can be estimated from partial updates computed by a subset of the workers. We apply the proposed scheme to the problem of tracking multiple vehicles. We show that replication achieves significantly higher accuracy than the corresponding uncoded scheme. The use of maximum distance separable (MDS) codes further improves accuracy for larger update intervals. In both cases, the proposed scheme approaches the accuracy of an ideal centralized scheme when the update interval is large enough. Finally, we observe a trade-off between age-of-information and estimation accuracy for MDS codes.

cs.IT↗

arXiv2018

Block-Diagonal and LT Codes for Distributed Computing With Straggling Servers

We propose two coded schemes for the distributed computing problem of multiplying a matrix by a set of vectors. The first scheme is based on partitioning the matrix into submatrices and applying maximum distance separable (MDS) codes to each submatrix. For this scheme, we prove that up to a given number of partitions the communication load and the computational delay (not including the encoding and decoding delay) are identical to those of the scheme recently proposed by Li et al., based on a single, long MDS code. However, due to the use of shorter MDS codes, our scheme yields a significantly lower overall computational delay when the delay incurred by encoding and decoding is also considered. We further propose a second coded scheme based on Luby Transform (LT) codes under inactivation decoding. Interestingly, LT codes may reduce the delay over the partitioned scheme at the expense of an increased communication load. We also consider distributed computing under a deadline and show numerically that the proposed schemes outperform other schemes in the literature, with the LT code-based scheme yielding the best performance for the scenarios considered.

cs.IT↗

arXiv2018

A Droplet Approach Based on Raptor Codes for Distributed Computing With Straggling Servers

We propose a coded distributed computing scheme based on Raptor codes to address the straggler problem. In particular, we consider a scheme where each server computes intermediate values, referred to as droplets, that are either stored locally or sent over the network. Once enough droplets are collected, the computation can be completed. Compared to previous schemes in the literature, our proposed scheme achieves lower computational delay when the decoding time is taken into account.

cs.IT↗

arXiv2017

Block-Diagonal Coding for Distributed Computing With Straggling Servers

We consider the distributed computing problem of multiplying a set of vectors with a matrix. For this scenario, Li et al. recently presented a unified coding framework and showed a fundamental tradeoff between computational delay and communication load. This coding framework is based on maximum distance separable (MDS) codes of code length proportional to the number of rows of the matrix, which can be very large. We propose a block-diagonal coding scheme consisting of partitioning the matrix into submatrices and encoding each submatrix using a shorter MDS code. We show that the assignment of coded matrix rows to servers to minimize the communication load can be formulated as an integer program with a nonlinear cost function, and propose an algorithm to solve it. We further prove that, up to a level of partitioning, the proposed scheme does not incur any loss in terms of computational delay (as defined by Li et al.) and communication load compared to the scheme by Li et al.. We also show numerically that, when the decoding time is also taken into account, the proposed scheme significantly lowers the overall computational delay with respect to the scheme by Li et al.. For heavy partitioning, this is achieved at the expense of a slight increase in communication load.

cs.IT↗