Nonlinearly weighted multiple kernel learning for time series forecasting

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

Abstract

Machine Learning methods such as Neural Network (NN) and Support Vector Regression (SVR) have been studied extensively for time series forecasting. Multiple Kernel Learning (MKL) which utilizes SVR as the predictor is yet another recent approaches to choose suitable kernels from a given pool of kernels by means of a linear combination of some base kernels. However, some literatures suggest that this linear combination of kernels cannot consistently outperform either the uniform combination of base kernels or simply the best single kernel. Hence, in this paper, other combination method is devised, namely the squared combination of base kernels, which gives more weight on suitable kernels and vice versa. We use time series data having various length, pattern and horizons, namely the 111 time series from NN3 competition, 3003 of M3 competition, 1001 of Ml competition and reduced 111 of Ml competition. Our experimental results indicate that our new forecasting approaches using squared combination of Multiple Kernel Learning (MKL) may perform well compared to the other methods on the same dataset.

Original languageEnglish
Title of host publicationProceedings - ICACSIS 2014
Subtitle of host publication2014 International Conference on Advanced Computer Science and Information Systems
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages390-395
Number of pages6
ISBN (Electronic)9781479980758
DOIs
Publication statusPublished - 23 Mar 2014
Event2014 International Conference on Advanced Computer Science and Information Systems, ICACSIS 2014 - Jakarta, Indonesia
Duration: 18 Oct 201419 Oct 2014

Publication series

NameProceedings - ICACSIS 2014: 2014 International Conference on Advanced Computer Science and Information Systems

Conference

Conference2014 International Conference on Advanced Computer Science and Information Systems, ICACSIS 2014
Country/TerritoryIndonesia
CityJakarta
Period18/10/1419/10/14

Fingerprint

Dive into the research topics of 'Nonlinearly weighted multiple kernel learning for time series forecasting'. Together they form a unique fingerprint.

Cite this