Extended features based random vector functional link network for classification problem

Random vector functional link (RVFL) network has been successfully employed in diverse domains such as computer vision and machine learning, due to its universal approximation capability. Recently, the shallow RVFL architecture has been extended to deep architectures. In deep architectures, multiple...

Full description

Saved in:
Bibliographic Details
Main Authors: Malik, Ashwani Kumar, Ganaie, M. A., Tanveer, M., Suganthan, Ponnuthurai Nagaratnam
Other Authors: School of Electrical and Electronic Engineering
Format: Article
Language:English
Published: 2022
Subjects:
Online Access:https://hdl.handle.net/10356/163627
Tags: Add Tag
No Tags, Be the first to tag this record!
Institution: Nanyang Technological University
Language: English
id sg-ntu-dr.10356-163627
record_format dspace
spelling sg-ntu-dr.10356-1636272022-12-13T01:43:48Z Extended features based random vector functional link network for classification problem Malik, Ashwani Kumar Ganaie, M. A. Tanveer, M. Suganthan, Ponnuthurai Nagaratnam School of Electrical and Electronic Engineering Engineering::Electrical and electronic engineering Deep Random Vector Functional Link Ensemble Learning Random vector functional link (RVFL) network has been successfully employed in diverse domains such as computer vision and machine learning, due to its universal approximation capability. Recently, the shallow RVFL architecture has been extended to deep architectures. In deep architectures, multiple hidden layers are stacked for extracting informative features from the original feature space. Therefore, having rich features, deep models are very successful compared to shallow models. In this article, we propose an extended feature RVFL (efRVFL) model that is trained over extended feature space generated analytically from the original feature space. The proposed efRVFL model has three types of features, i.e., original features, supervised randomized (newly generated) features, and unsupervised randomized features, in its feature matrix. The proposed efRVFL model with additional features has capability to capture nonlinear hidden relationships within the dataset. The proposed efRVFL model is an unstable classifier, and thus, its performance can be improved further via ensemble learning. Ensemble models are stable and accurate and have better generalization performance than single models. Therefore, we also propose an ensemble of extended feature RVFL (en-efRVFL) model. Each base model of en-efRVFL is trained over different feature spaces so that more accurate and diverse base models can be generated. The outcome of the base models is integrated via average voting scheme. Empirical evaluation over <inline-formula> <tex-math notation="LaTeX">$46$</tex-math> </inline-formula> UCI classification datasets demonstrates that the proposed efRVFL and en-efRVFL models have better performance than RVFL and other given deep models. Furthermore, the experimental results over <inline-formula> <tex-math notation="LaTeX">$12$</tex-math> </inline-formula> sparse datasets show that the proposed en-efRVFL model has a winning performance among several deep feedforward neural networks (FNNs). This work was supported in part by the National Supercomputing Mission under the Department of Science and Technology (DST) and Miety, Government of India, under Grant DST/NSM/R&D HPC_Appl/2021/03.29; in part by the Department of Science and Technology through the Interdisciplinary Cyber Physical Systems (ICPS) Scheme under Grant DST/ICPS/CPSIndividual/2018/276; in part by the Science and Engineering Research Board (SERB), Government of India, through the Ramanujan Fellowship Scheme, under Grant SB/S2/RJN-001/2016; and in part by the Council of Scientific and Industrial Research (CSIR), New Delhi, India, through the Extra Mural Research (EMR) Scheme under Grant 22(0751)/17/EMR-II. The work of Ashwani Kumar Malik was supported by the CSIR, New Delhi, India, under Grant 09/1022 (0075)/2019-EMR-I. 2022-12-13T01:43:48Z 2022-12-13T01:43:48Z 2022 Journal Article Malik, A. K., Ganaie, M. A., Tanveer, M. & Suganthan, P. N. (2022). Extended features based random vector functional link network for classification problem. IEEE Transactions On Computational Social Systems, 1-10. https://dx.doi.org/10.1109/TCSS.2022.3187461 2329-924X https://hdl.handle.net/10356/163627 10.1109/TCSS.2022.3187461 2-s2.0-85136114424 1 10 en IEEE Transactions on Computational Social Systems © 2022 IEEE. All rights reserved.
institution Nanyang Technological University
building NTU Library
continent Asia
country Singapore
Singapore
content_provider NTU Library
collection DR-NTU
language English
topic Engineering::Electrical and electronic engineering
Deep Random Vector Functional Link
Ensemble Learning
spellingShingle Engineering::Electrical and electronic engineering
Deep Random Vector Functional Link
Ensemble Learning
Malik, Ashwani Kumar
Ganaie, M. A.
Tanveer, M.
Suganthan, Ponnuthurai Nagaratnam
Extended features based random vector functional link network for classification problem
description Random vector functional link (RVFL) network has been successfully employed in diverse domains such as computer vision and machine learning, due to its universal approximation capability. Recently, the shallow RVFL architecture has been extended to deep architectures. In deep architectures, multiple hidden layers are stacked for extracting informative features from the original feature space. Therefore, having rich features, deep models are very successful compared to shallow models. In this article, we propose an extended feature RVFL (efRVFL) model that is trained over extended feature space generated analytically from the original feature space. The proposed efRVFL model has three types of features, i.e., original features, supervised randomized (newly generated) features, and unsupervised randomized features, in its feature matrix. The proposed efRVFL model with additional features has capability to capture nonlinear hidden relationships within the dataset. The proposed efRVFL model is an unstable classifier, and thus, its performance can be improved further via ensemble learning. Ensemble models are stable and accurate and have better generalization performance than single models. Therefore, we also propose an ensemble of extended feature RVFL (en-efRVFL) model. Each base model of en-efRVFL is trained over different feature spaces so that more accurate and diverse base models can be generated. The outcome of the base models is integrated via average voting scheme. Empirical evaluation over <inline-formula> <tex-math notation="LaTeX">$46$</tex-math> </inline-formula> UCI classification datasets demonstrates that the proposed efRVFL and en-efRVFL models have better performance than RVFL and other given deep models. Furthermore, the experimental results over <inline-formula> <tex-math notation="LaTeX">$12$</tex-math> </inline-formula> sparse datasets show that the proposed en-efRVFL model has a winning performance among several deep feedforward neural networks (FNNs).
author2 School of Electrical and Electronic Engineering
author_facet School of Electrical and Electronic Engineering
Malik, Ashwani Kumar
Ganaie, M. A.
Tanveer, M.
Suganthan, Ponnuthurai Nagaratnam
format Article
author Malik, Ashwani Kumar
Ganaie, M. A.
Tanveer, M.
Suganthan, Ponnuthurai Nagaratnam
author_sort Malik, Ashwani Kumar
title Extended features based random vector functional link network for classification problem
title_short Extended features based random vector functional link network for classification problem
title_full Extended features based random vector functional link network for classification problem
title_fullStr Extended features based random vector functional link network for classification problem
title_full_unstemmed Extended features based random vector functional link network for classification problem
title_sort extended features based random vector functional link network for classification problem
publishDate 2022
url https://hdl.handle.net/10356/163627
_version_ 1753801177031507968