{"type":"ScholarlyArticle","title":"Modelling the Working Week for Multi-Step Forecasting using Gaussian Process Regression","authors":[{"name":"Pasan Karunaratne","slug":"pasan-karunaratne","given":"Pasan","family":"Karunaratne","affiliation":"Department of Computing and Information Systems , The University of Melbourne , Australia"},{"name":"Masud Moshtaghi","slug":"masud-moshtaghi","given":"Masud","family":"Moshtaghi","affiliation":"Department of Computing and Information Systems , The University of Melbourne , Australia"},{"name":"Shanika Karunasekera","slug":"shanika-karunasekera","given":"Shanika","family":"Karunasekera","affiliation":"Department of Computing and Information Systems , The University of Melbourne , Australia"},{"name":"Aaron Harwood","slug":"aaron-harwood","given":"Aaron","family":"Harwood","affiliation":"Department of Computing and Information Systems , The University of Melbourne , Australia"},{"name":"Trevor Cohn","slug":"trevor-cohn","given":"Trevor","family":"Cohn","affiliation":"Department of Computing and Information Systems , The University of Melbourne , Australia"}],"abstract":"In time-series forecasting, regression is a popular method, with Gaussian Process Regression widely held to be the state of the art. The versatility of Gaussian Processes has led to them being used in many varied application domains. However, though many real-world applications involve data which follows a working-week structure, where weekends exhibit substantially different behavior to weekdays, methods for explicit modelling of working-week effects in Gaussian Process Regression models have not been proposed. Not explicitly modelling the working week fails to incorporate a significant source of information which can be invaluable in forecasting scenarios. In this work we provide novel kernel-combination methods to explicitly model working-week effects in time-series data for more accurate predictions using Gaussian Process Regression. Further, we demonstrate that prediction accuracy can be improved by constraining the non-convex optimization process of finding optimal hyperparameter values. We validate the effectiveness of our methods by performing multi-step prediction on two real-world publicly available time-series datasets - one relating to electricity Smart Meter data of the University of Melbourne, and the other relating to the counts of pedestrians in the City of Melbourne.","doi":"10.24963/ijcai.2017/277","url":"https://ijcai.org/proceedings/2017/277","pdf":"https://ijcai.org/proceedings/2017/0277.pdf","markdown":"https://ijcai.org/proceedings/2017/277.md","bibtex":"https://ijcai.org/proceedings/2017/bibtex/277","volume":{"slug":"2017","title":"Proceedings of the Twenty-Sixth International Joint Conference on Artificial Intelligence","short_title":"IJCAI 2017","url":"https://ijcai.org/proceedings/2017/","year":2017,"location":"Melbourne, Australia","dates":"19-25 August 2017","editors":"Carles Sierra","isbn":"978-0-9992411-0-3","publisher":"International Joint Conferences on Artificial Intelligence Organization","doi_prefix":"10.24963/ijcai.2017"},"section":"Machine Learning A-R","track":"Main track","number":277,"pages":{"first":1994,"last":2000},"keywords":["Machine Learning: Time-series/Data Streams","Machine Learning: Machine Learning"],"videos":[],"references":[{"position":1,"raw":"[Alzate and Sinn, 2013] Carlos Alzate and Mathieu Sinn. Improved electricity load forecasting via kernel spectral clustering of smart meters. In 2013 IEEE 13th Interna- tional Conference on Data Mining, pages 943-948. IEEE, 2013.","title":"Improved electricity load forecasting via kernel spectral clustering of smart meters","authors":"Alzate, Sinn, Carlos Alzate, Mathieu Sinn","year":"2013","venue":"IEEE 13th International Conference on Data Mining","doi":null,"url":null,"internal_url":null},{"position":2,"raw":"[Bickel et al., 2008] Steffen Bickel, Jasmina Bogojeska, Thomas Lengauer, and Tobias Scheffer. Multi-task learn- ing for hiv therapy screening. In Proceedings of the 25th international conference on Machine learning, pages 56- 63. ACM, 2008.","title":"Multi-task learning for hiv therapy screening","authors":"Steffen Bickel, Jasmina Bogojeska, Thomas Lengauer, Tobias Scheffer","year":"2008","venue":"Proceedings of the 25th international conference on Machine learning","doi":null,"url":null,"internal_url":null},{"position":3,"raw":"[Chen et al., 2013] Niya Chen, Zheng Qian, Ian T Nabney, and Xiaofeng Meng. Short-term wind power forecasting using gaussian processes. In IJCAI, 2013.","title":"Short-term wind power forecasting using gaussian processes","authors":"Niya Chen, Zheng Qian, Ian T Nabney, Xiaofeng Meng","year":"2013","venue":"IJCAI","doi":null,"url":null,"internal_url":"/proceedings/2013/411"},{"position":4,"raw":"[Conejo et al., 2005] Antonio J Conejo, Miguel A Plazas, Rosa Espinola, and Ana B Molina. Day-ahead electric- ity price forecasting using the wavelet transform and arima models. IEEE transactions on power systems, 20(2):1035- 1042, 2005.","title":"Day-ahead electricity price forecasting using the wavelet transform and arima models","authors":"Miguel A Antonio J Conejo, Rosa Plazas, Ana B Espinola, Molina","year":"2005","venue":"IEEE transactions on power systems","doi":null,"url":null,"internal_url":null},{"position":5,"raw":"[Doan et al., 2015] Minh Tuan Doan, Sutharshan Ra- jasegarar, Mahsa Salehi, Masud Moshtaghi, and Christo- pher Leckie. Profiling pedestrian distribution and anomaly detection in a dynamic environment. In Proceedings of the 24th ACM International on Conference on Information and Knowledge Management, pages 1827-1830. ACM, 2015.","title":"Profiling pedestrian distribution and anomaly detection in a dynamic environment","authors":"Minh Tuan Doan, Sutharshan Rajasegarar, Mahsa Salehi, Masud Moshtaghi, Christopher Leckie","year":"2015","venue":"Proceedings of the 24th ACM International on Conference on Information and Knowledge Management","doi":null,"url":null,"internal_url":null},{"position":6,"raw":"[Guan et al., 2013] Che Guan, Peter B Luh, Laurent D Michel, Yuting Wang, and Peter B Friedland. Very short-term load forecasting: wavelet neural networks with data pre-filtering. IEEE Transactions on Power Systems, 28(1):30-41, 2013.","title":"Very short-term load forecasting: wavelet neural networks with data pre-filtering","authors":"Che Guan, Laurent D Peter B Luh, Yuting Michel, Peter B Wang, Friedland","year":"2013","venue":"IEEE Transactions on Power Systems","doi":null,"url":null,"internal_url":null},{"position":7,"raw":"[Hachino and Kadirkamanathan, 2007] Tomohiro Hachino and Visakan Kadirkamanathan. Time series forecasting using multiple gaussian process prior model. In Compu- tational Intelligence and Data Mining, 2007. CIDM 2007. IEEE Symposium on, pages 604-609. IEEE, 2007.","title":"Time series forecasting using multiple gaussian process prior model","authors":"Tomohiro Hachino, Visakan Kadirkamanathan","year":"2007","venue":"Computational Intelligence and Data Mining","doi":null,"url":null,"internal_url":null},{"position":8,"raw":"[Klenske et al., 2013] Edgar D Klenske, Melanie N Zeilinger, Bernhard Schölkopf, and Philipp Hennig. Nonparametric dynamics estimation for time periodic systems. In Communication, Control, and Computing (Allerton), 2013 51st Annual Allerton Conference on, pages 486-493. IEEE, 2013.","title":"Nonparametric dynamics estimation for time periodic systems","authors":"Melanie N Edgar D Klenske, Bernhard Zeilinger, Philipp Schölkopf, Hennig","year":"2013","venue":"Communication, Control, and Computing (Allerton)","doi":null,"url":null,"internal_url":null},{"position":9,"raw":"[Kolter and Ferreira, 2011] J Zico Kolter and Joseph Fer- reira. A large-scale study on predicting and contextual- izing building energy usage. In Twenty-Fifth AAAI Con- ference on Artificial Intelligence, 2011.","title":"A large-scale study on predicting and contextualizing building energy usage","authors":"Ferreira ; J Zico Kolter, Joseph Kolter, Ferreira","year":"2011","venue":"Twenty-Fifth AAAI Conference on Artificial Intelligence","doi":null,"url":null,"internal_url":null},{"position":10,"raw":"[Krajnik et al., 2014] Tomas Krajnik, Jaime Pulido Fen- tanes, Grzegorz Cielniak, Christian Dondrup, and Tom Duckett. Spectral analysis for long-term robotic mapping. In Robotics and Automation (ICRA), 2014 IEEE Interna- tional Conference on, pages 3706-3711. IEEE, 2014.","title":"Spectral analysis for long-term robotic mapping","authors":"Krajnik","year":"2014","venue":"Robotics and Automation (ICRA), 2014 IEEE International Conference on","doi":null,"url":null,"internal_url":null},{"position":11,"raw":"[Leith et al., 2004] Douglas J Leith, Martin Heidl, and John Ringwood. Gaussian process prior models for electrical load forecasting. Probabilistic Methods Applied to Power Systems, pages 112-117, 2004.","title":"Gaussian process prior models for electrical load forecasting","authors":"Martin Douglas J Leith, John Heidl, Ringwood","year":"2004","venue":"Probabilistic Methods Applied to Power Systems","doi":null,"url":null,"internal_url":null},{"position":12,"raw":"[Preotiuc-Pietro and Cohn, 2013] Daniel Preotiuc-Pietro and Trevor Cohn. A temporal model of text periodicities using gaussian processes. In EMNLP, pages 977-988, 2013.","title":"A temporal model of text periodicities using gaussian processes","authors":"Preotiuc-Pietro, Cohn, Daniel Preotiuc-Pietro, Trevor Cohn","year":"2013","venue":"EMNLP","doi":null,"url":null,"internal_url":null},{"position":13,"raw":"[Rasmussen, 2006] Carl Edward Rasmussen. Gaussian pro- cesses for machine learning. 2006.","title":"Gaussian processes for machine learning","authors":"Carl Edward, Rasmussen","year":"2006","venue":null,"doi":null,"url":null,"internal_url":null},{"position":14,"raw":"[Senanayake et al., 2016] Ransalu Senanayake, Simon O'Callaghan, and Fabio Ramos. Predicting spatio- temporal propagation of seasonal influenza using variational gaussian process regression. In Thirtieth AAAI Conference on Artificial Intelligence, 2016.","title":"Predicting spatiotemporal propagation of seasonal influenza using variational gaussian process regression","authors":"Senanayake","year":"2016","venue":"Thirtieth AAAI Conference on Artificial Intelligence","doi":null,"url":null,"internal_url":null},{"position":15,"raw":"[Sturm and Burgard, 2013] Jürgen Sturm and Wolfram Bur- gard. Learning probabilistic models for mobile manipula- tion robots. In IJCAI, 2013.","title":"Jürgen Sturm and Wolfram Burgard. Learning probabilistic models for mobile manipulation robots","authors":"Burgard Sturm","year":"2013","venue":"IJCAI","doi":null,"url":null,"internal_url":null},{"position":16,"raw":"[Williams et al., 2009] Christopher Williams, Stefan Klanke, Sethu Vijayakumar, and Kian M Chai. Multi-task gaussian process learning of robot inverse dynamics. In Advances in Neural Information Processing Systems, pages 265-272, 2009.","title":"Multi-task gaussian process learning of robot inverse dynamics","authors":"Christopher Williams, Stefan Klanke, Sethu Vijayakumar, Kian M Chai","year":"2009","venue":"Advances in Neural Information Processing Systems","doi":null,"url":null,"internal_url":null}],"license_note":"Copyright International Joint Conferences on Artificial Intelligence."}