PSO based Neural Networks vs. traditional statistical models for seasonal time series forecasting

Seasonality is a distinctive characteristic which is often observed in many practical time series. Artificial Neural Networks (ANNs) are a class of promising models for efficiently recognizing and forecasting seasonal patterns. In this paper, the Particle Swarm Optimization (PSO) approach is used to...

Ausführliche Beschreibung

Gespeichert in:
Bibliographische Detailangaben
Hauptverfasser: Adhikari, R., Agrawal, R. K., Kant, L.
Format: Tagungsbericht
Sprache:eng
Schlagworte:
Online-Zugang:Volltext bestellen
Tags: Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
container_end_page 725
container_issue
container_start_page 719
container_title
container_volume
creator Adhikari, R.
Agrawal, R. K.
Kant, L.
description Seasonality is a distinctive characteristic which is often observed in many practical time series. Artificial Neural Networks (ANNs) are a class of promising models for efficiently recognizing and forecasting seasonal patterns. In this paper, the Particle Swarm Optimization (PSO) approach is used to enhance the forecasting strengths of feedforward ANN (FANN) as well as Elman ANN (EANN) models for seasonal data. Three widely popular versions of the basic PSO algorithm, viz. Trelea-I, Trelea-II and Clerc-Type1 are considered here. The empirical analysis is conducted on three real-world seasonal time series. Results clearly show that each version of the PSO algorithm achieves notably better forecasting accuracies than the standard Backpropagation (BP) training method for both FANN and EANN models. The neural network forecasting results are also compared with those from the three traditional statistical models, viz. Seasonal Autoregressive Integrated Moving Average (SARIMA), Holt-Winters (HW) and Support Vector Machine (SVM). The comparison demonstrates that both PSO and BP based neural networks outperform SARIMA, HW and SVM models for all three time series datasets. The forecasting performances of ANNs are further improved through combining the outputs from the three PSO based models.
doi_str_mv 10.1109/IAdCC.2013.6514315
format Conference Proceeding
fullrecord <record><control><sourceid>ieee_6IE</sourceid><recordid>TN_cdi_ieee_primary_6514315</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>6514315</ieee_id><sourcerecordid>6514315</sourcerecordid><originalsourceid>FETCH-LOGICAL-i90t-620e0757a29c3915bfd5eaae6090c9915ad61e2d253ee029d938a6107c1cfda93</originalsourceid><addsrcrecordid>eNpVkM1OwzAQhI0QEqjkBeCSF0hY27FTH6uIn0oVRaL3amtvkCFpkG1AvD1W2wun2dlvdg7L2A2HmnMwd8uF67paAJe1VryRXJ2xwrRz3uhWNkoYef7Pt-aSFTG-A0C-11zDFcOX13W5w0iufKavgEOW9DOFj1h-x7pMAZ1PftpnEBMmH5O3eR4nR0Ms-ymUkTAeePIjZRc8HQBZzOH92zW76HGIVJx0xjYP95vuqVqtH5fdYlV5A6nSAgha1aIwVhqudr1ThEgaDFiTF-g0J-GEkkQgjDNyjppDa7ntHRo5Y7fHWk9E28_gRwy_29Nf5B-ygleX</addsrcrecordid><sourcetype>Publisher</sourcetype><iscdi>true</iscdi><recordtype>conference_proceeding</recordtype></control><display><type>conference_proceeding</type><title>PSO based Neural Networks vs. traditional statistical models for seasonal time series forecasting</title><source>IEEE Electronic Library (IEL) Conference Proceedings</source><creator>Adhikari, R. ; Agrawal, R. K. ; Kant, L.</creator><creatorcontrib>Adhikari, R. ; Agrawal, R. K. ; Kant, L.</creatorcontrib><description>Seasonality is a distinctive characteristic which is often observed in many practical time series. Artificial Neural Networks (ANNs) are a class of promising models for efficiently recognizing and forecasting seasonal patterns. In this paper, the Particle Swarm Optimization (PSO) approach is used to enhance the forecasting strengths of feedforward ANN (FANN) as well as Elman ANN (EANN) models for seasonal data. Three widely popular versions of the basic PSO algorithm, viz. Trelea-I, Trelea-II and Clerc-Type1 are considered here. The empirical analysis is conducted on three real-world seasonal time series. Results clearly show that each version of the PSO algorithm achieves notably better forecasting accuracies than the standard Backpropagation (BP) training method for both FANN and EANN models. The neural network forecasting results are also compared with those from the three traditional statistical models, viz. Seasonal Autoregressive Integrated Moving Average (SARIMA), Holt-Winters (HW) and Support Vector Machine (SVM). The comparison demonstrates that both PSO and BP based neural networks outperform SARIMA, HW and SVM models for all three time series datasets. The forecasting performances of ANNs are further improved through combining the outputs from the three PSO based models.</description><identifier>ISBN: 9781467345279</identifier><identifier>ISBN: 146734527X</identifier><identifier>EISBN: 9781467345293</identifier><identifier>EISBN: 1467345288</identifier><identifier>EISBN: 9781467345286</identifier><identifier>EISBN: 1467345296</identifier><identifier>DOI: 10.1109/IAdCC.2013.6514315</identifier><language>eng</language><publisher>IEEE</publisher><subject>ANN ; Artificial neural networks ; Box-Jenkins models ; Computational modeling ; Elman ANN ; Forecasting ; particle swarm optimization ; Predictive models ; seasonality ; Support vector machines ; Time series analysis ; time series forecasting ; Training</subject><ispartof>2013 3rd IEEE International Advance Computing Conference (IACC), 2013, p.719-725</ispartof><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/6514315$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>309,310,776,780,785,786,2052,27902,54895</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/6514315$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Adhikari, R.</creatorcontrib><creatorcontrib>Agrawal, R. K.</creatorcontrib><creatorcontrib>Kant, L.</creatorcontrib><title>PSO based Neural Networks vs. traditional statistical models for seasonal time series forecasting</title><title>2013 3rd IEEE International Advance Computing Conference (IACC)</title><addtitle>IAdCC</addtitle><description>Seasonality is a distinctive characteristic which is often observed in many practical time series. Artificial Neural Networks (ANNs) are a class of promising models for efficiently recognizing and forecasting seasonal patterns. In this paper, the Particle Swarm Optimization (PSO) approach is used to enhance the forecasting strengths of feedforward ANN (FANN) as well as Elman ANN (EANN) models for seasonal data. Three widely popular versions of the basic PSO algorithm, viz. Trelea-I, Trelea-II and Clerc-Type1 are considered here. The empirical analysis is conducted on three real-world seasonal time series. Results clearly show that each version of the PSO algorithm achieves notably better forecasting accuracies than the standard Backpropagation (BP) training method for both FANN and EANN models. The neural network forecasting results are also compared with those from the three traditional statistical models, viz. Seasonal Autoregressive Integrated Moving Average (SARIMA), Holt-Winters (HW) and Support Vector Machine (SVM). The comparison demonstrates that both PSO and BP based neural networks outperform SARIMA, HW and SVM models for all three time series datasets. The forecasting performances of ANNs are further improved through combining the outputs from the three PSO based models.</description><subject>ANN</subject><subject>Artificial neural networks</subject><subject>Box-Jenkins models</subject><subject>Computational modeling</subject><subject>Elman ANN</subject><subject>Forecasting</subject><subject>particle swarm optimization</subject><subject>Predictive models</subject><subject>seasonality</subject><subject>Support vector machines</subject><subject>Time series analysis</subject><subject>time series forecasting</subject><subject>Training</subject><isbn>9781467345279</isbn><isbn>146734527X</isbn><isbn>9781467345293</isbn><isbn>1467345288</isbn><isbn>9781467345286</isbn><isbn>1467345296</isbn><fulltext>true</fulltext><rsrctype>conference_proceeding</rsrctype><creationdate>2013</creationdate><recordtype>conference_proceeding</recordtype><sourceid>6IE</sourceid><sourceid>RIE</sourceid><recordid>eNpVkM1OwzAQhI0QEqjkBeCSF0hY27FTH6uIn0oVRaL3amtvkCFpkG1AvD1W2wun2dlvdg7L2A2HmnMwd8uF67paAJe1VryRXJ2xwrRz3uhWNkoYef7Pt-aSFTG-A0C-11zDFcOX13W5w0iufKavgEOW9DOFj1h-x7pMAZ1PftpnEBMmH5O3eR4nR0Ms-ymUkTAeePIjZRc8HQBZzOH92zW76HGIVJx0xjYP95vuqVqtH5fdYlV5A6nSAgha1aIwVhqudr1ThEgaDFiTF-g0J-GEkkQgjDNyjppDa7ntHRo5Y7fHWk9E28_gRwy_29Nf5B-ygleX</recordid><startdate>201302</startdate><enddate>201302</enddate><creator>Adhikari, R.</creator><creator>Agrawal, R. K.</creator><creator>Kant, L.</creator><general>IEEE</general><scope>6IE</scope><scope>6IL</scope><scope>CBEJK</scope><scope>RIE</scope><scope>RIL</scope></search><sort><creationdate>201302</creationdate><title>PSO based Neural Networks vs. traditional statistical models for seasonal time series forecasting</title><author>Adhikari, R. ; Agrawal, R. K. ; Kant, L.</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-i90t-620e0757a29c3915bfd5eaae6090c9915ad61e2d253ee029d938a6107c1cfda93</frbrgroupid><rsrctype>conference_proceedings</rsrctype><prefilter>conference_proceedings</prefilter><language>eng</language><creationdate>2013</creationdate><topic>ANN</topic><topic>Artificial neural networks</topic><topic>Box-Jenkins models</topic><topic>Computational modeling</topic><topic>Elman ANN</topic><topic>Forecasting</topic><topic>particle swarm optimization</topic><topic>Predictive models</topic><topic>seasonality</topic><topic>Support vector machines</topic><topic>Time series analysis</topic><topic>time series forecasting</topic><topic>Training</topic><toplevel>online_resources</toplevel><creatorcontrib>Adhikari, R.</creatorcontrib><creatorcontrib>Agrawal, R. K.</creatorcontrib><creatorcontrib>Kant, L.</creatorcontrib><collection>IEEE Electronic Library (IEL) Conference Proceedings</collection><collection>IEEE Proceedings Order Plan All Online (POP All Online) 1998-present by volume</collection><collection>IEEE Xplore All Conference Proceedings</collection><collection>IEEE Electronic Library (IEL)</collection><collection>IEEE Proceedings Order Plans (POP All) 1998-Present</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Adhikari, R.</au><au>Agrawal, R. K.</au><au>Kant, L.</au><format>book</format><genre>proceeding</genre><ristype>CONF</ristype><atitle>PSO based Neural Networks vs. traditional statistical models for seasonal time series forecasting</atitle><btitle>2013 3rd IEEE International Advance Computing Conference (IACC)</btitle><stitle>IAdCC</stitle><date>2013-02</date><risdate>2013</risdate><spage>719</spage><epage>725</epage><pages>719-725</pages><isbn>9781467345279</isbn><isbn>146734527X</isbn><eisbn>9781467345293</eisbn><eisbn>1467345288</eisbn><eisbn>9781467345286</eisbn><eisbn>1467345296</eisbn><abstract>Seasonality is a distinctive characteristic which is often observed in many practical time series. Artificial Neural Networks (ANNs) are a class of promising models for efficiently recognizing and forecasting seasonal patterns. In this paper, the Particle Swarm Optimization (PSO) approach is used to enhance the forecasting strengths of feedforward ANN (FANN) as well as Elman ANN (EANN) models for seasonal data. Three widely popular versions of the basic PSO algorithm, viz. Trelea-I, Trelea-II and Clerc-Type1 are considered here. The empirical analysis is conducted on three real-world seasonal time series. Results clearly show that each version of the PSO algorithm achieves notably better forecasting accuracies than the standard Backpropagation (BP) training method for both FANN and EANN models. The neural network forecasting results are also compared with those from the three traditional statistical models, viz. Seasonal Autoregressive Integrated Moving Average (SARIMA), Holt-Winters (HW) and Support Vector Machine (SVM). The comparison demonstrates that both PSO and BP based neural networks outperform SARIMA, HW and SVM models for all three time series datasets. The forecasting performances of ANNs are further improved through combining the outputs from the three PSO based models.</abstract><pub>IEEE</pub><doi>10.1109/IAdCC.2013.6514315</doi><tpages>7</tpages></addata></record>
fulltext fulltext_linktorsrc
identifier ISBN: 9781467345279
ispartof 2013 3rd IEEE International Advance Computing Conference (IACC), 2013, p.719-725
issn
language eng
recordid cdi_ieee_primary_6514315
source IEEE Electronic Library (IEL) Conference Proceedings
subjects ANN
Artificial neural networks
Box-Jenkins models
Computational modeling
Elman ANN
Forecasting
particle swarm optimization
Predictive models
seasonality
Support vector machines
Time series analysis
time series forecasting
Training
title PSO based Neural Networks vs. traditional statistical models for seasonal time series forecasting
url https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-16T23%3A42%3A12IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-ieee_6IE&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=proceeding&rft.atitle=PSO%20based%20Neural%20Networks%20vs.%20traditional%20statistical%20models%20for%20seasonal%20time%20series%20forecasting&rft.btitle=2013%203rd%20IEEE%20International%20Advance%20Computing%20Conference%20(IACC)&rft.au=Adhikari,%20R.&rft.date=2013-02&rft.spage=719&rft.epage=725&rft.pages=719-725&rft.isbn=9781467345279&rft.isbn_list=146734527X&rft_id=info:doi/10.1109/IAdCC.2013.6514315&rft_dat=%3Cieee_6IE%3E6514315%3C/ieee_6IE%3E%3Curl%3E%3C/url%3E&rft.eisbn=9781467345293&rft.eisbn_list=1467345288&rft.eisbn_list=9781467345286&rft.eisbn_list=1467345296&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rft_ieee_id=6514315&rfr_iscdi=true