ApproxML: Efficient approximate AdHoc ML models through materialization and reuse

Sona Hasani, Faezeh Ghaderi, Shohedul Hasan, Saravanan Thirumuruganathan, Abolfazl Asudeh, Nick Koudasz, Gautam Das

Research output: Contribution to journalConference article

Abstract

Machine learning (ML) has gained a pivotal role in answering complex predictive analytic queries. Model building for large scale datasets is one of the time consuming parts of the data science pipeline. Often data scientists are willing to sacrfice some accuracy in order to speed up this process during the exploratory phase. In this paper, we propose to demonstrate ApproxML, a system that efficiently constructs approximate ML models for new queries from previously constructed ML models using the concepts of model materialization and reuse. ApproxML supports a variety of ML models such as generalized linear models for supervised learning, and K-means and Gaussian Mixture model for unsupervised learning.

Original languageEnglish
Pages (from-to)1906-1909
Number of pages4
JournalProceedings of the VLDB Endowment
Volume12
Issue number12
DOIs
Publication statusPublished - 1 Jan 2018
Event45th International Conference on Very Large Data Bases, VLDB 2019 - Los Angeles, United States
Duration: 26 Aug 201730 Aug 2017

Fingerprint

Learning systems
Unsupervised learning
Supervised learning
Pipelines

ASJC Scopus subject areas

  • Computer Science (miscellaneous)
  • Computer Science(all)

Cite this

ApproxML : Efficient approximate AdHoc ML models through materialization and reuse. / Hasani, Sona; Ghaderi, Faezeh; Hasan, Shohedul; Thirumuruganathan, Saravanan; Asudeh, Abolfazl; Koudasz, Nick; Das, Gautam.

In: Proceedings of the VLDB Endowment, Vol. 12, No. 12, 01.01.2018, p. 1906-1909.

Research output: Contribution to journalConference article

Hasani, Sona ; Ghaderi, Faezeh ; Hasan, Shohedul ; Thirumuruganathan, Saravanan ; Asudeh, Abolfazl ; Koudasz, Nick ; Das, Gautam. / ApproxML : Efficient approximate AdHoc ML models through materialization and reuse. In: Proceedings of the VLDB Endowment. 2018 ; Vol. 12, No. 12. pp. 1906-1909.
@article{03cfe56d8a5c45fea109e0a46af0f0db,
title = "ApproxML: Efficient approximate AdHoc ML models through materialization and reuse",
abstract = "Machine learning (ML) has gained a pivotal role in answering complex predictive analytic queries. Model building for large scale datasets is one of the time consuming parts of the data science pipeline. Often data scientists are willing to sacrfice some accuracy in order to speed up this process during the exploratory phase. In this paper, we propose to demonstrate ApproxML, a system that efficiently constructs approximate ML models for new queries from previously constructed ML models using the concepts of model materialization and reuse. ApproxML supports a variety of ML models such as generalized linear models for supervised learning, and K-means and Gaussian Mixture model for unsupervised learning.",
author = "Sona Hasani and Faezeh Ghaderi and Shohedul Hasan and Saravanan Thirumuruganathan and Abolfazl Asudeh and Nick Koudasz and Gautam Das",
year = "2018",
month = "1",
day = "1",
doi = "10.14778/3352063.3352096",
language = "English",
volume = "12",
pages = "1906--1909",
journal = "Proceedings of the VLDB Endowment",
issn = "2150-8097",
publisher = "Very Large Data Base Endowment Inc.",
number = "12",

}

TY - JOUR

T1 - ApproxML

T2 - Efficient approximate AdHoc ML models through materialization and reuse

AU - Hasani, Sona

AU - Ghaderi, Faezeh

AU - Hasan, Shohedul

AU - Thirumuruganathan, Saravanan

AU - Asudeh, Abolfazl

AU - Koudasz, Nick

AU - Das, Gautam

PY - 2018/1/1

Y1 - 2018/1/1

N2 - Machine learning (ML) has gained a pivotal role in answering complex predictive analytic queries. Model building for large scale datasets is one of the time consuming parts of the data science pipeline. Often data scientists are willing to sacrfice some accuracy in order to speed up this process during the exploratory phase. In this paper, we propose to demonstrate ApproxML, a system that efficiently constructs approximate ML models for new queries from previously constructed ML models using the concepts of model materialization and reuse. ApproxML supports a variety of ML models such as generalized linear models for supervised learning, and K-means and Gaussian Mixture model for unsupervised learning.

AB - Machine learning (ML) has gained a pivotal role in answering complex predictive analytic queries. Model building for large scale datasets is one of the time consuming parts of the data science pipeline. Often data scientists are willing to sacrfice some accuracy in order to speed up this process during the exploratory phase. In this paper, we propose to demonstrate ApproxML, a system that efficiently constructs approximate ML models for new queries from previously constructed ML models using the concepts of model materialization and reuse. ApproxML supports a variety of ML models such as generalized linear models for supervised learning, and K-means and Gaussian Mixture model for unsupervised learning.

UR - http://www.scopus.com/inward/record.url?scp=85074523659&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85074523659&partnerID=8YFLogxK

U2 - 10.14778/3352063.3352096

DO - 10.14778/3352063.3352096

M3 - Conference article

AN - SCOPUS:85074523659

VL - 12

SP - 1906

EP - 1909

JO - Proceedings of the VLDB Endowment

JF - Proceedings of the VLDB Endowment

SN - 2150-8097

IS - 12

ER -