Combining Deep Learning with Good Old-Fashioned Machine Learning

Moshe Sipper

doi:10.1007/s42979-022-01505-2

Combining Deep Learning with Good Old-Fashioned Machine Learning

Moshe Sipper

Research output: Contribution to journal › Article › peer-review

1 Scopus citations

Abstract

We present a comprehensive, stacking-based framework for combining deep learning with good old-fashioned machine learning, called Deep GOld. Our framework involves ensemble selection from 51 retrained pretrained deep networks as first-level models, and 10 machine-learning algorithms as second-level models. Enabled by today’s state-of-the-art software tools and hardware platforms, Deep GOld delivers consistent improvement when tested on four image-classification datasets: Fashion MNIST, CIFAR10, CIFAR100, and Tiny ImageNet. Of 120 experiments, in all but 10 Deep GOld improved the original networks’ performance.

Original language	English
Article number	85
Journal	SN Computer Science
Volume	4
Issue number	1
DOIs	https://doi.org/10.1007/s42979-022-01505-2
State	Published - 1 Jan 2023

Keywords

Deep learning
Image analysis
Machine learning
Pattern recognition

ASJC Scopus subject areas

General Computer Science
Computer Science Applications
Computer Networks and Communications
Computer Graphics and Computer-Aided Design
Computational Theory and Mathematics
Artificial Intelligence

Access to Document

10.1007/s42979-022-01505-2

Cite this

@article{8758b691e70046da9cdfd261059f74f4,

title = "Combining Deep Learning with Good Old-Fashioned Machine Learning",

abstract = "We present a comprehensive, stacking-based framework for combining deep learning with good old-fashioned machine learning, called Deep GOld. Our framework involves ensemble selection from 51 retrained pretrained deep networks as first-level models, and 10 machine-learning algorithms as second-level models. Enabled by today{\textquoteright}s state-of-the-art software tools and hardware platforms, Deep GOld delivers consistent improvement when tested on four image-classification datasets: Fashion MNIST, CIFAR10, CIFAR100, and Tiny ImageNet. Of 120 experiments, in all but 10 Deep GOld improved the original networks{\textquoteright} performance.",

keywords = "Deep learning, Image analysis, Machine learning, Pattern recognition",

author = "Moshe Sipper",

note = "Publisher Copyright: {\textcopyright} 2022, The Author(s), under exclusive licence to Springer Nature Singapore Pte Ltd.",

year = "2023",

month = jan,

day = "1",

doi = "10.1007/s42979-022-01505-2",

language = "English",

volume = "4",

journal = "SN Computer Science",

issn = "2662-995X",

publisher = "Springer",

number = "1",

}

TY - JOUR

T1 - Combining Deep Learning with Good Old-Fashioned Machine Learning

AU - Sipper, Moshe

PY - 2023/1/1

Y1 - 2023/1/1

N2 - We present a comprehensive, stacking-based framework for combining deep learning with good old-fashioned machine learning, called Deep GOld. Our framework involves ensemble selection from 51 retrained pretrained deep networks as first-level models, and 10 machine-learning algorithms as second-level models. Enabled by today’s state-of-the-art software tools and hardware platforms, Deep GOld delivers consistent improvement when tested on four image-classification datasets: Fashion MNIST, CIFAR10, CIFAR100, and Tiny ImageNet. Of 120 experiments, in all but 10 Deep GOld improved the original networks’ performance.

AB - We present a comprehensive, stacking-based framework for combining deep learning with good old-fashioned machine learning, called Deep GOld. Our framework involves ensemble selection from 51 retrained pretrained deep networks as first-level models, and 10 machine-learning algorithms as second-level models. Enabled by today’s state-of-the-art software tools and hardware platforms, Deep GOld delivers consistent improvement when tested on four image-classification datasets: Fashion MNIST, CIFAR10, CIFAR100, and Tiny ImageNet. Of 120 experiments, in all but 10 Deep GOld improved the original networks’ performance.

KW - Deep learning

KW - Image analysis

KW - Machine learning

KW - Pattern recognition

UR - https://www.scopus.com/pages/publications/85143623548

U2 - 10.1007/s42979-022-01505-2

DO - 10.1007/s42979-022-01505-2

M3 - Article

AN - SCOPUS:85143623548

SN - 2662-995X

VL - 4

JO - SN Computer Science

JF - SN Computer Science

IS - 1

M1 - 85

ER -

Combining Deep Learning with Good Old-Fashioned Machine Learning

Abstract

Keywords

ASJC Scopus subject areas

Access to Document

Other files and links

Fingerprint

Cite this