ds-array: A Distributed Data Structure for Large Scale Machine Learning

TL;DR

This paper introduces ds-array, a new distributed data structure for dislib that significantly improves performance, scalability, and usability for large-scale scientific machine learning tasks on HPC clusters.

Contribution

The paper presents ds-array, a novel distributed data structure that overcomes dislib's limitations by offering a NumPy-like API and enhanced efficiency.

Findings

Performance improvements of up to 100x over Datasets

Enhanced scalability and usability in scientific data analysis

Reduced computational complexity of key operations

Abstract

Machine learning has proved to be a useful tool for extracting knowledge from scientific data in numerous research fields, including astrophysics, genomics, and molecular dynamics. Often, data sets from these research areas need to be processed in distributed platforms due to their magnitude. This can be done using one of the various distributed machine learning libraries available. One of these libraries is dislib, a distributed machine learning library for Python especially designed to process large scale data sets on HPC clusters, which makes dislib an ideal candidate for analyzing scientific data. However, dislib's main distributed data structure, called Dataset, has some limitations, including poor performance in certain operations and low flexibility and usability. In this paper, we propose a novel distributed data structure for dislib, called ds-array, that addresses dislib's main limitations in data management. Ds-arrays simplify distributed data management in dislib by exposing a NumPy-like API, provide more flexibility, and reduce the computational complexity of some operations. This results in performance improvements of up to two orders of magnitude over Datasets, while also greatly improving scalability and usability.