個人資料
|
水谷英二 | Eiji Mizutani
|
- 最高學歷:國立清華大學資訊工程博士 (Ph.D in Computer Science)
- 教授課程:動態規劃、排程、馬可夫決策過程、機率模型、非線性最佳化
- 研究領域:動態規劃、數值線性代數、非線性最佳化、機器學習法
- 實驗室:非線性數值最佳化實驗室(MA011-1) 分機:7937
|
各位新進學生:
我這邊有一個國科會資助國內研究生的研究機會, 內容是與斯圖爾特德雷福斯教授(加州大學柏克萊分校之名譽教授)共同研究的一個小型計畫. 在研究所學習期間, 可有柏克萊大學之旅的補助. 如果你有興趣請用電子郵件並用英文書寫來跟我聯絡.
Selected recent papers (已挑選)
- Eiji Mizutani and Stuart Dreyfus."Totally model-free actor-critic recurrent neural-network reinforcement learning in non-Markovian domains." Annals of Operations Research(SCI) ,pp 1-25,2016. DOI: 10.1007/s10479-016-2366-2
- Eiji Mizutani. "On Pantoja's problem allegedly showing a distinction between differential dynamic programming and stagewise Newton methods." International Journal of Control, Taylor & Francis (SCI), vol.8, issue 9, pp.1702-1711, 2015.
- Eiji Mizutani. "A counterexample to a proposed dynamic programming algorithm for optimal bid construction in an auction-based fully distributed manufacturing system." International Journal of Advanced Manufacturing Technology, Springer (SCI), Vol.71, pp.377-380, 2014.
- Eiji Mizutani. "A note on dynamic programming formulations for scheduling job classes with changeover times on a single machine." In Proceedings of the IEEE International Conference on Industrial Engineering and Engineering Management, (Received an Outstanding Paper Award), five pages, (EI), 2013.
- Eiji Mizutani. "Reformulation of Lawler's algorithm by auxiliary-information dynamic programming in a minimax-cost scheduling problem." In Proceedings of the IEEE International Conference on Industrial Engineering and Engineering Management, pp.192-196, (EI), 2012.
- Eiji Mizutani and James Demmel. "On improving trust-region variable projection algorithms for separable nonlinear least squares learning." In Proceedings of IJCNN 2011 (EI), pp.397-404.
- Eiji Mizutani and Stuart Dreyfus. "An analysis on negative curvature induced by singularity in multi-layer neural-network learning." Advances in Neural Information Processing Systems, Vol. 23, pages1669-1677, 2010.
- Eiji Mizutani and Stuart Dreyfus. "Stage-lookahead dynamic programming algorithms for stochastic problems with time-lagged control dynamics." In Proceedings of the IEEE International Conference on Industrial Engineering and Engineering Management, pp.301-305 (EI), 2009.
- Eiji Mizutani and Stuart Dreyfus. "Second-order stagewise backpropagation for Hessian-matrix analyses and investigation of negative curvature." Journal of Neural Networks (SCI), Elsevier Science, vol.21, pp.193-203,2008 (Errata).
- Eiji Mizutani and Jing-Yun Carey Fan. "On exploiting symmetry for multilayer perceptron learning." In Proceedings of IJCNN, pp.2857-2862, 2007.
- Eiji Mizutani and Stuart E. Dreyfus. "On derivation of stagewise second-order backpropagation by invariant imbedding for multi-stage neural-network learning." In Proceedings of the INNS-IEEE International Joint Conference on Neural Networks, Vancouver, CANADA, July 2006.
- Eiji Mizutani and Stuart E. Dreyfus. "Stagewise Newton, differential dynamic programming, and neighboring optimum control for neural-network learning." In Proceedings of the American Control Conference (ACC'05), pp. 1331-1336, Portland Oregon, June 8-10, 2005.
- Eiji Mizutani, Stuart E. Dreyfus, and James W. Demmel. "Second-order backpropagation algorithms for a stagewise-partitioned separable Hessian matrix." In Proceedings of the INNS-IEEE International Joint Conference on Neural Networks, Montreal CANADA, August 2005.
- Eiji Mizutani and James W. Demmel. "Iterative scaled trust-region learning in Krylov subspaces via Pearlmutter's implicit sparse Hessian-vector multiply." Advances in Neural Information Processing Systems, Vol. 16, pages 209-216, edited by S. Thrun, L. Saul, and B. Scholkopf, MIT Press, 2004.
- Eiji Mizutani and Stuart E. Dreyfus. "On using discretized Cohen-Grossberg node dynamics for model-free actor-critic neural learning in non-Markovian domains." In Proceedings of the 5th IEEE International Symposium on Computational Intelligence in Robotics and Automation (CIRA2003), Vol.1, pages 1-6, Kobe JAPAN, July 16-20, 2003.
- Eiji Mizutani and James W. Demmel. "On structure-exploiting trust-region regularized nonlinear least squares algorithms for neural-network learning." Journal of Neural Networks (SCI), Elsevier Science, Vol. 16, pp. 745-753, 2003.
- Eiji Mizutani and Kenichi Nishio. "Multi-Illuminant Color Reproduction for Electronic Cameras Via CANFIS Neuro-Fuzzy Modular Network Device Characterization." IEEE Transactions on Neural Networks (SCI), Vol. 13, No. 4, pages 1009-1022, July 2002.
- Eiji Mizutani and James W. Demmel. "On separable nonlinear least squares algorithms for neuro-fuzzy modular network learning." In Proceedings of the INNS-IEEE International Joint Conference on Neural Networks, Vol.3, pp. 2399-2404, Honolulu USA, May 2002.
- Eiji Mizutani and James W. Demmel. "On Iterative Krylov-Dogleg Trust-Region Steps for Solving Neural Networks Nonlinear Least Squares Problems." Advances in Neural Information Processing Systems, vol.13, pp. 605-611, MIT Press, 2001.
- Eiji Mizutani, Stuart Dreyfus, and Ken-ichi Nishio. "On derivation of MLP backpropagation from the Kelley-Bryson optimal-control gradient formula and its application." In Proceedings of the IEEE-INNS International Joint Conference on Neural Networks, pp. 2167-2172 (vol.2), Como, Italy, July, 2000.
- Eiji Mizutani. "Sample path-based policy-only learning by actor neural networks." In Proceedings of the IEEE International Conference on Neural Networks, pp. 1245-1250 (vol.2) , 1999.
其他著作 (Book)
- J.-S. R. Jang, C.-T. Sun, and E. Mizutani. Neuro-Fuzzy and Soft Computing: a computational approach to learning and machine intelligence, Prentice Hall, Upper Saddle River, NJ, 1997 (614 pages).
國際專利
- Sony Corporation: Ken-ichi Nishio and Eiji Mizutani. "Methods and apparatus for color device characterization." PCT WO 0065841.
- Matsushita Electric Co., Kansai Paint Co., Ltd, and University of California, Berkeley (Takagi, Mizutani, and Auslander). "Proportion predicting systems and methods of making mixture." US 6081796A.
更新日期:2016/12/01