LearnedSort as a learning-augmented SampleSort: Analysis and Parallelization

TL;DR

This paper analyzes and parallelizes LearnedSort, a machine learning-based sorting algorithm, demonstrating its effectiveness as a learning-augmented SampleSort with improved parallel performance over existing algorithms.

Contribution

It provides a formal analysis of LearnedSort as a learning-augmented SampleSort and develops a parallel version combining it with IPS4o, enhancing sorting efficiency.

Findings

Parallel LearnedSort outperforms IPS4o on synthetic datasets.

Parallel LearnedSort shows improved performance on real-world datasets.

Analysis confirms LearnedSort's theoretical basis as a learning-augmented SampleSort.

Abstract

This work analyzes and parallelizes LearnedSort, the novel algorithm that sorts using machine learning models based on the cumulative distribution function. LearnedSort is analyzed under the lens of algorithms with predictions, and it is argued that LearnedSort is a learning-augmented SampleSort. A parallel LearnedSort algorithm is developed combining LearnedSort with the state-of-the-art SampleSort implementation, IPS4o. Benchmarks on synthetic and real-world datasets demonstrate improved parallel performance for parallel LearnedSort compared to IPS4o and other sorting algorithms.