New approaches for the analysis ofcomplex time-series using kernel methods.

• Project/Area Number: 23700172
• Research Category: Grant-in-Aid for Young Scientists (B)
• Allocation Type: Multi-year Fund
• Research Field: Intelligent informatics
• Research Institution: Kyoto University
• Principal Investigator: CUTURI Marco 京都大学, 情報学研究科, 准教授 (80597344)
• Project Period (FY): 2011 – 2012
• Project Status: Completed (Fiscal Year 2012)
• Budget Amount: ¥4,290,000 (Direct Cost: ¥3,300,000、Indirect Cost: ¥990,000); Fiscal Year 2012: ¥1,820,000 (Direct Cost: ¥1,400,000、Indirect Cost: ¥420,000); Fiscal Year 2011: ¥2,470,000 (Direct Cost: ¥1,900,000、Indirect Cost: ¥570,000)
• Keywords: Machine Learning / Kernel Methods / 時列カーネル / 国際研究者交流

Research Abstract

Time series are now increasingly complex. Each observation may describe a structured object (an image or a graph for instance) or alternatively a very high dimensional feature vector. The goal of our project is to develop new methods to handle time-series of complex data through kernel methods and optimization methods.

Research Products

• [Journal Article] Mean Reversion with a Variance Threshold (2013)
  • Author(s): M. Cuturi, A. d'Aspremont
  • Journal Title: JMLR W&CP Volume: 28(3) Pages: 271-279
• [Journal Article] Autoregressive Kernels for Time Series (2013)
  • Author(s): Marco Cuturi, Arnaud Doucet
  • Journal Title: Journal of Machine Learning Research Volume: pending
  • Peer Reviewed
• [Journal Article] Mean Reversion with a Variance Threshold (2013)
  • Author(s): Marco Cuturi, Alexandre d'Aspremont
  • Journal Title: International Conference on Machine Learning 2013 Volume: 2013 Proceedings/Journal
  • Peer Reviewed
• [Journal Article] Autoregressive Kernels for Time Series
  • Author(s): M. Cuturi, A.Doucet
  • Journal Title: Journal of Machine Learning Research
• [Presentation] Kernel Methods for Time Series (2012)
  • Author(s): Marco Cuturi
  • Organizer: 29th International Conference on Machine Learning (ICML), Kernel Methods Workshop(招待講演)
  • Place of Presentation: Edinburgh, UK
• [Presentation] Fast Global Alignment Kernels (2011)
  • Author(s): Marco Cuturi
  • Organizer: 28th International Conference on Machine Learning (ICML)
  • Place of Presentation: Bellevue, Washington, USA
• [Presentation] Fast Global Alignment Kernels
  • Author(s): M. Cuturi
  • Organizer: Proceedings of the International Conference on Machine Learning 2011
• [Presentation] Transportation Distances and their Application in Machine Learning: New Problems
  • Author(s): Marco Cuturi
  • Organizer: Cambridge Machine Learning Seminar
  • Place of Presentation: Massachusetts Institute of Technology (USA)
  • Invited
• [Presentation] Transportation Distances and their Application in Machine Learning: New Problems
  • Author(s): Marco Cuturi
  • Organizer: University of California Los Angeles Electrical Engineering Seminar
  • Place of Presentation: University of California Los Angeles (USA)
• [Presentation] Distances and Kernels on Discrete Structures: the generating-function trick
  • Author(s): Marco Cuturi
  • Organizer: Institut National de la Recherche en Informatique et Automatique, LEAR Seminar
  • Place of Presentation: Institut National de la Recherche en Informatique et Automatique Grenoble (France)
  • Invited
• [Presentation] Transportation and Machine Learning
  • Author(s): Marco Cuturi
  • Organizer: Statistical Machine Learning in Paris Seminar
  • Place of Presentation: Ecole Nationale Superieure des Mines de Paris (France)
  • Invited
• [Presentation] Distances and Kernels on Discrete Structures
  • Author(s): Marco Cuturi
  • Organizer: Sapporo Workshop on Machine Learning and Applications to Biology
  • Place of Presentation: Hokkaido University
• [Presentation] Distances and Kernels on Discrete Structures
  • Author(s): Marco Cuturi
  • Organizer: International Conference on Machine Learning Workshop on RKHS and Kernel Based Methods
  • Place of Presentation: University of Edinburgh (UK)
  • Invited
• [Remarks]
  • URL: http://www.iip.ist.i.kyoto-u.ac.jp/member/cuturi/GA.html
• [Remarks]
  • URL: http://www.iip.ist.i.kyoto-u.ac.jp/member/cuturi/AR.html
• [Remarks] Triangular Global Alignment Kernels
  • URL: http://www.iip.ist.i.kyoto-u.ac.jp/member/cuturi/GA.html
• [Remarks] Autoregressive Kernels
  • URL: http://www.iip.ist.i.kyoto-u.ac.jp/member/cuturi/AR.html