RankMap: A Framework for Distributed Learning From Dense Data Sets

Azalia Mirhoseini

doi:10.1109/TNNLS.2016.2631581

RankMap: A Framework for Distributed Learning From Dense Data Sets

2018-07-01

3

OA

AI

DOI：10.1109/TNNLS.2016.2631581

Original

Save

View PDF

Abstract

En 中文

This paper introduces RankMap, a platform-aware end-to-end framework for efficient execution of a broad class of iterative learning algorithms for massive and dense data sets. Our framework exploits data structure to scalably factorize it into an ensemble of lower rank subspaces. The factorization creates sparse low-dimensional representations of the data, a property which is leveraged to devise effective mapping and scheduling of iterative learning algorithms on the distributed computing machines. We provide two APIs, one matrix-based and one graph-based, which facilitate automated adoption of the framework for performing several contemporary learning applications. To demonstrate the utility of RankMap, we solve sparse recovery and power iteration problems on various real-world data sets with up to 1.8 billion nonzeros. Our evaluations are performed on Amazon EC2 and IBM iDataPlex servers using up to 244 cores. The results demonstrate up to two orders of magnitude improvements in memory usage, execution speed, and bandwidth compared with the best reported prior work, while achieving the same level of learning accuracy.

Keywords:

Dense and big data

iterative machine learning (ML)

large-scale distributed computing

sparse matrix factorization

low rank approximation

union of subspaces

RankMap: A Framework for Distributed Learning From Dense Data Sets

Abstract

AI Summary

Journal

Organization