A Machine Learning Pattern for CNNs with Mixed-Feature Modes

Aklil Zenebe Kiflay; Raimund Kirner; Athanasios Tsokanos

doi:10.18745/pb.25867

A Machine Learning Pattern for CNNs with Mixed-Feature Modes

Aklil Zenebe Kiflay, Raimund Kirner, Athanasios Tsokanos

Research output: Chapter in Book/Report/Conference proceeding › Conference contribution

Abstract

Convolutional Neural Network (CNN) is a Deep Neural Network (DNN) that is well-suited to local feature extraction. DNN learns features useful for a particular learning problem from raw input data, thus reducing reliance on domain experts. However, automatic learning of features via DNN comes at a cost, as it requires relatively high computing resources. An approach to reduce DNN’s resource requirement is to use additional hand-crafted features as input, providing a compromise between automation and resource demand. In this paper, we present a hybrid DNN with mixed-mode features, and demonstrate its application in intrusion detection using UNSWNB15 dataset.

Original language	English
Title of host publication	Proceedings of Abstracts, School of Physics, Engineering and Computer Science Research Conference 2022
Publisher	University of Hertfordshire
Pages	75-77
DOIs	https://doi.org/10.18745/pb.25867
Publication status	Published - 4 Nov 2022
Event	2nd School of Physics, Engineering and Computer Science Research Conference - Online - hosted by the School of Physics, Engineering and Computer Science, Hatfield, United Kingdom Duration: 12 Apr 2022 → 12 Apr 2022 https://uhra.herts.ac.uk/handle/2299/25867

Conference

Conference	2nd School of Physics, Engineering and Computer Science Research Conference
Abbreviated title	SPECS 2022
Country/Territory	United Kingdom
City	Hatfield
Period	12/04/22 → 12/04/22
Internet address	https://uhra.herts.ac.uk/handle/2299/25867

Access to Document

10.18745/pb.25867Licence: CC BY

Cite this

@inproceedings{680b397b52064354be9944ef66d8c19e,

title = "A Machine Learning Pattern for CNNs with Mixed-Feature Modes",

abstract = "Convolutional Neural Network (CNN) is a Deep Neural Network (DNN) that is well-suited to local feature extraction. DNN learns features useful for a particular learning problem from raw input data, thus reducing reliance on domain experts. However, automatic learning of features via DNN comes at a cost, as it requires relatively high computing resources. An approach to reduce DNN{\textquoteright}s resource requirement is to use additional hand-crafted features as input, providing a compromise between automation and resource demand. In this paper, we present a hybrid DNN with mixed-mode features, and demonstrate its application in intrusion detection using UNSWNB15 dataset.",

author = "Kiflay, \{Aklil Zenebe\} and Raimund Kirner and Athanasios Tsokanos",

year = "2022",

month = nov,

day = "4",

doi = "10.18745/pb.25867",

language = "English",

pages = "75--77",

booktitle = "Proceedings of Abstracts, School of Physics, Engineering and Computer Science Research Conference 2022",

publisher = "University of Hertfordshire",

note = "2nd School of Physics, Engineering and Computer Science Research Conference, SPECS 2022 ; Conference date: 12-04-2022 Through 12-04-2022",

url = "https://uhra.herts.ac.uk/handle/2299/25867",

}

Kiflay, AZ, Kirner, R & Tsokanos, A 2022, A Machine Learning Pattern for CNNs with Mixed-Feature Modes. in Proceedings of Abstracts, School of Physics, Engineering and Computer Science Research Conference 2022. University of Hertfordshire, pp. 75-77, 2nd School of Physics, Engineering and Computer Science Research Conference, Hatfield, United Kingdom, 12/04/22. https://doi.org/10.18745/pb.25867

TY - GEN

T1 - A Machine Learning Pattern for CNNs with Mixed-Feature Modes

AU - Kiflay, Aklil Zenebe

AU - Kirner, Raimund

AU - Tsokanos, Athanasios

PY - 2022/11/4

Y1 - 2022/11/4

N2 - Convolutional Neural Network (CNN) is a Deep Neural Network (DNN) that is well-suited to local feature extraction. DNN learns features useful for a particular learning problem from raw input data, thus reducing reliance on domain experts. However, automatic learning of features via DNN comes at a cost, as it requires relatively high computing resources. An approach to reduce DNN’s resource requirement is to use additional hand-crafted features as input, providing a compromise between automation and resource demand. In this paper, we present a hybrid DNN with mixed-mode features, and demonstrate its application in intrusion detection using UNSWNB15 dataset.

AB - Convolutional Neural Network (CNN) is a Deep Neural Network (DNN) that is well-suited to local feature extraction. DNN learns features useful for a particular learning problem from raw input data, thus reducing reliance on domain experts. However, automatic learning of features via DNN comes at a cost, as it requires relatively high computing resources. An approach to reduce DNN’s resource requirement is to use additional hand-crafted features as input, providing a compromise between automation and resource demand. In this paper, we present a hybrid DNN with mixed-mode features, and demonstrate its application in intrusion detection using UNSWNB15 dataset.

U2 - 10.18745/pb.25867

DO - 10.18745/pb.25867

M3 - Conference contribution

SP - 75

EP - 77

BT - Proceedings of Abstracts, School of Physics, Engineering and Computer Science Research Conference 2022

PB - University of Hertfordshire

T2 - 2nd School of Physics, Engineering and Computer Science Research Conference

Y2 - 12 April 2022 through 12 April 2022

ER -

A Machine Learning Pattern for CNNs with Mixed-Feature Modes

Abstract

Conference

Access to Document

Fingerprint

Cite this