Combination of Feature Selection and Learning Methods for IoT Data Fusion

In this paper, we propose five data fusion schemes for the Internet of Things (IoT) scenario,which are Relief and Perceptron (Re-P), Relief and Genetic Algorithm Particle Swarm Optimization (Re-GAPSO), Genetic Algorithm and Artificial Neural Network (GA-ANN), Rough and Perceptron (Ro-P)and Rough and...

Full description

Saved in:
Bibliographic Details
Main Authors: V. Sattari-Naeini, Zahra Parizi-Nejad
Format: Article
Language:English
Published: Amirkabir University of Technology 2017-12-01
Series:AUT Journal of Electrical Engineering
Subjects:
Online Access:https://eej.aut.ac.ir/article_1960_5b7511e4f87d3b6a9eb1a6bc95cececc.pdf
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849420390904889344
author V. Sattari-Naeini
Zahra Parizi-Nejad
author_facet V. Sattari-Naeini
Zahra Parizi-Nejad
author_sort V. Sattari-Naeini
collection DOAJ
description In this paper, we propose five data fusion schemes for the Internet of Things (IoT) scenario,which are Relief and Perceptron (Re-P), Relief and Genetic Algorithm Particle Swarm Optimization (Re-GAPSO), Genetic Algorithm and Artificial Neural Network (GA-ANN), Rough and Perceptron (Ro-P)and Rough and GAPSO (Ro-GAPSO). All the schemes consist of four stages, including preprocessingthe data set based on curve fitting, reducing the data dimension and identifying the most effective featuresets according to data correlation, training classification algorithms, and finally predicting new databased on classification algorithms. The results derived from five compound schemes are investigated andcompared with each other with three metrics, namely, Quality of Train (QoT) Accuracy (Ac) and StorageCapacity (SC). While the Re-P scheme is only capable of separating classes that are linearly separable,Re-GAPSO one is a dynamic method, appropriate for constantly changing problems of the real life. Onthe other hand, GA-ANN is a Wrapper method and despite Relief can adapt itself to the machine learningalgorithm. Meanwhile, Ro-P scheme is useful for analyzing vague and imprecise information and, unlikeGA-ANN, has less calculative costs. Among these five schemes, Ro-GAPSO is a more precise one, whichhas less calculative cost and does not become stuck in local minima. Experimental results show that Re-Poutperforms other proposed and existing methods in terms of computational time complexity.
format Article
id doaj-art-b46fb5bec7704c5884048a21f66e3dcd
institution Kabale University
issn 2588-2910
2588-2929
language English
publishDate 2017-12-01
publisher Amirkabir University of Technology
record_format Article
series AUT Journal of Electrical Engineering
spelling doaj-art-b46fb5bec7704c5884048a21f66e3dcd2025-08-20T03:31:46ZengAmirkabir University of TechnologyAUT Journal of Electrical Engineering2588-29102588-29292017-12-0149222323210.22060/eej.2017.12151.50461960Combination of Feature Selection and Learning Methods for IoT Data FusionV. Sattari-Naeini0Zahra Parizi-Nejad1Dept. of Computer Engineering, Shahid Bahonar University of Kerman, Kerman, IranDept. of Computer Engineering, Shahid Bahonar University of Kerman, Kerman, IranIn this paper, we propose five data fusion schemes for the Internet of Things (IoT) scenario,which are Relief and Perceptron (Re-P), Relief and Genetic Algorithm Particle Swarm Optimization (Re-GAPSO), Genetic Algorithm and Artificial Neural Network (GA-ANN), Rough and Perceptron (Ro-P)and Rough and GAPSO (Ro-GAPSO). All the schemes consist of four stages, including preprocessingthe data set based on curve fitting, reducing the data dimension and identifying the most effective featuresets according to data correlation, training classification algorithms, and finally predicting new databased on classification algorithms. The results derived from five compound schemes are investigated andcompared with each other with three metrics, namely, Quality of Train (QoT) Accuracy (Ac) and StorageCapacity (SC). While the Re-P scheme is only capable of separating classes that are linearly separable,Re-GAPSO one is a dynamic method, appropriate for constantly changing problems of the real life. Onthe other hand, GA-ANN is a Wrapper method and despite Relief can adapt itself to the machine learningalgorithm. Meanwhile, Ro-P scheme is useful for analyzing vague and imprecise information and, unlikeGA-ANN, has less calculative costs. Among these five schemes, Ro-GAPSO is a more precise one, whichhas less calculative cost and does not become stuck in local minima. Experimental results show that Re-Poutperforms other proposed and existing methods in terms of computational time complexity.https://eej.aut.ac.ir/article_1960_5b7511e4f87d3b6a9eb1a6bc95cececc.pdfinternet of thingsdata fusionrough set theoryperceptrongapso
spellingShingle V. Sattari-Naeini
Zahra Parizi-Nejad
Combination of Feature Selection and Learning Methods for IoT Data Fusion
AUT Journal of Electrical Engineering
internet of things
data fusion
rough set theory
perceptron
gapso
title Combination of Feature Selection and Learning Methods for IoT Data Fusion
title_full Combination of Feature Selection and Learning Methods for IoT Data Fusion
title_fullStr Combination of Feature Selection and Learning Methods for IoT Data Fusion
title_full_unstemmed Combination of Feature Selection and Learning Methods for IoT Data Fusion
title_short Combination of Feature Selection and Learning Methods for IoT Data Fusion
title_sort combination of feature selection and learning methods for iot data fusion
topic internet of things
data fusion
rough set theory
perceptron
gapso
url https://eej.aut.ac.ir/article_1960_5b7511e4f87d3b6a9eb1a6bc95cececc.pdf
work_keys_str_mv AT vsattarinaeini combinationoffeatureselectionandlearningmethodsforiotdatafusion
AT zahraparizinejad combinationoffeatureselectionandlearningmethodsforiotdatafusion