Hidekazu Oiwa

Email: hidekazu.oiwa at gmail.com

Software Engineer (Machine Learning) at Google


Ph.D (Information Science and Technology) at the University of Tokyo, Japan.
My research interests are Machine Learning and Natural Language Processing.
Japanese Page is here.

Professional

Recruit Holdings Co., Ltd. 2015/04 >> 2017/12

  • Researcher

Stanford University 2015/09 >> 2017/08

  • Visiting Scholar

NEC Laboratories America

  • Visiting Research
  • Supervisor: Dr. Ryohei Fujimaki

Microsoft Research Asia

  • Research Internship
  • Supervisor: Dr. Junichi Tsujii

Preferred Infrastructure

  • Internship
Education

Ph.D., Mathematical Informatics, The University of Tokyo 2010/04 >> 2015/03

  • A receiptant of Research Fellowship for Young Scientists (DC1) from JSPS
  • Concentration: Machine Learning, Natural Language Processing
  • Supervisor: Hiroshi Nakagawa

BA, Engineering, The University of Tokyo 2006/04 >> 2010/03

  • Concentration: Financial Statistics, Economics
  • Supervisor: Fumiko Takeda
Selected Publications

Conference (refereed)

  • "Knowledge Transfer for Out-of-Knowledge-Base Entities: A Graph Neural Network Approach"
    • Takuo Hamaguchi, Hidekazu Oiwa, Masashi Shinbo, Yuji Matsumoto
    • International Joint Conference on Artificial Intelligence. (IJCAI 2017)
  • "Partition-wise Linear Models"
    • Hidekazu Oiwa, Ryohei Fujimaki
    • Advances in Neural Information Processing Systems (NIPS 2014)
    • [Paper]
  • "Formalizing Word Sampling for Vocabulary Prediction as Graph-based Active Learning"
    • Yo Ehara, Yusuke Miyao, Hidekazu Oiwa, Issei Sato, Hiroshi Nakagawa
    • Conference on Empirical Methods in Natural Language Processing (EMNLP 2014)
  • "Robust Distributed Training of Linear Classifiers Based on Divergence Minimization Principle"
    • Junpei Komiyama, Hidekazu Oiwa, Hiroshi Nakagawa
    • European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML/PKDD 2014)
  • "Common Space Embedding of Primal-Dual Relation Semantic Spaces"
    • Hidekazu Oiwa, Jun'ichi Tsujii
    • International Conference on Computational Linguistics (COLING-2014)
  • "Online and Stochastic Learning with a Human Cognitive Bias"
  • "Mining words in the minds of second language learners: learner-specific word difficulty"
    • Yo Ehara, Issei Sato, Hidekazu Oiwa, Hiroshi Nakagawa
    • International Conference on Computational Linguistics (COLING-2012)
  • "Healing Truncation Bias : Self-weighted Truncation framework for Dual Averaging"
    • Hidekazu Oiwa, Shin Matsushima, Hiroshi Nakagawa
    • IEEE International Conference on Data Mining (ICDM 2012)
    • [Preprint] [Slides]
  • "Frequency-aware Truncated methods for Sparse Online Learning"
    • Hidekazu Oiwa, Shin Matsushima, Hiroshi Nakagawa
    • European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML/PKDD 2011)
  • "The Economic Impact of Herd Behavior in the Japanese Loan Market"
    • Ryuichi Nakagawa, Hidekazu Oiwa, Fumiko Takeda
    • Western Economic Association International Annual Conference (WEAI 2010)

Journal (refereed)

  • "Feature-aware Regularization for Sparse Online Learning"
    • Hidekazu Oiwa, Shin Matsushima, Hiroshi Nakagawa
    • Science China Information Sciences, Vol. 57 Issue 5 1-21, 2014.
  • "The Economic Impact of Herd Behavior in the Japanese Loan Market"
    • Ryuichi Nakagawa, Hidekazu Oiwa, Fumiko Takeda
    • Pacific-Basin Finance Journal, Vol.20 Issue 4 600-613, 2012.
  • "L1 regularized online supervised learning using feature frequency"
    • Hidekazu Oiwa, Shin Matsushima, Hiroshi Nakagawa
    • The Information Processing Society of Japan (IPSJ):Transactions on Mathematical Modeling and its Applications (TOM), Vol.4 No.3 84-93, 2011.

Workshop

  • "Novel Sparse Modeling by L2 + L0 Regularization”
    • Hidekazu Oiwa, Issei Sato, Hiroshi Nakagawa
    • NIPS Workshop on Discrete and Combinatorial Problems in Machine Learning (DISCML 2013)
  • "Understanding seed selection in bootstrapping"
    • Yo Ehara, Issei Sato, Hidekazu Oiwa, and Hiroshi Nakagawa
    • Graph-Based Methods for Natural Language Processing (TextGraphs-8), 2013
Honors / Achievements

Microsoft Research Asia Excellent Internship Award2014

  • Microsoft Research Asia

JSAI Student Incentive AwardJun, 2013

  • The 27th Annual Conference of the Japanese Society for Artificial Intelligence

IPSJ Computer Science Research Award for Young Scientists

Presentation AwardMar, 2011

  • 82th Conference of Mathematical Modeling and Problem Solving.
  • Paper: "L1 regularized online supervised learning using feature frequency."

Enginnering Dean's Award for students

Outstanding Research AwardMar, 2010

  • The University of Tokyo.
  • Paper: "The Economic Impact of Herd Behavior in the Japanese Loan Market."
Programming Skills
  • C++
  • Python
  • JavaScript
  • Github (including online learning classifier library)