MULP: A Multi-Layer Perceptron Application to Long-Term, Out-of-Sample Time Series Prediction

Item Type: Proceeding
MIUR type: Proceedings > Proceedings
Title: MULP: A Multi-Layer Perceptron Application to Long-Term, Out-of-Sample Time Series Prediction
Authors string: Pasero E; Raimondo G; Ruffa S.
University authors:
Page Range: pp. 566-575
Publisher: Springer
ISBN: 9783642133176
ISSN: 0302-9743
Volume: 6064
Event Title: 7th International Symposium on Neural Networks, ISNN 2010
Event Location: Shanghai
Event Dates: 6-9 June 2010
Abstract: A forecasting approach based on Multi-Layer Perceptron (MLP) Artificial Neural Networks (named by the authors MULP) is proposed for the NN5 111 time series long-term, out of sample forecasting competition. This approach follows a direct prediction strategy and is completely automatic. It has been chosen after having been compared with other regression methods (as for example Support Vector Machines (SVMs)) and with a recursive approach to prediction. Good results have also been obtained using the ANNs forecaster together with a dimensional reduction of the input features space performed through a Principal Component Analysis (PCA) and a proper information theory based backward selection algorithm. Using this methodology we took the 10th place among the best 50% scorers in the final results table of the NN5 competition
Date: 2010
Status: Published
Language of publication: English
Uncontrolled Keywords: time series analysis, artificial neural networks, machine learning
Departments (original): DISPEA - Production Systems and Business Economics
DELEN - Electronics
Departments: DET - Department of Electronics and Telecommunications
Related URLs:
    Subjects: Area 09 - Ingegneria industriale e dell'informazione > ELETTRONICA
    Date Deposited: 07 Jun 2010 12:44
    Last modification data (IRIS): 18 May 2015 11:00:04
    Update date (PORTO): 19 May 2015 00:10
    Id Number (DOI): 10.1007/978-3-642-13318-3_70
    Linksolver URL: Linksolver link

    This field presents the citations number present on Scopus and Web of Science databases e links to the remote records. Also Google Scholar link is present.

    There may be discrepancies with respect to the data in databases for the following reasons:

    • Differences from fields (title, year,...) in UGOV and those in the databases.
    • PORTO citations are extracted monthly. The db is in real time
    • The WoS citation number reflect the collections subscribed by Politecnico (Science citation index Expanded and Conference Proceedings Citation Index)

    For informations contact scrivia/porto



    [img] PDF (Multilayerperceptr_2010) - Postprint
    Document access: Not visible (accessible only to the record owner)
    Licence: Not public - Private access / Restricted.

    Download (232Kb) | Send a request to the author for a copy of the paper

    Actions (login required)

    View Item View Item