Hisashi Kashima has been a research staff member of Data Analytics Group in Tokyo Research Laboratory of IBM Research since April, 1999.
He is working on machine learning and data mining research and their applications to bioinformatics, autonomic computing, and business intelligence, and also contributes to academic societies.
He obtained his B.S. degree in applied mathematics and physics in 1997, and a M.S. degree in systems engineering in 1999, and a Ph.D. degree in informatics in 2007 from Kyoto University in Japan.

Current Research Interests: Machine Learning

• Design and analysis of learning algorithms for structured data
  • Kernel methods for structured data
    • Efficient and expressive kernel functions for structured data such as sequences, trees, and graphs
      • Tree kernels, graph kernels, ...
  • Structure mapping learning
    • Learning algorithms for inferring mapping from structured data to structured data
    • Kernel methods for structure mapping
  • Network structure prediction
    • Learning algorithms for inferring structure of networks such as biological networks and social networks
  • Optimization problems accompanied by machine learning
• Applications of machine learning techniques to ...
  • Problem detection and analysis for computer systems
  • Bioinformatics
  • Natural language processing
  • Marketing science
  • Financial engineering
  • Social network analysis

Awards

• 2007 Winner, IEEE Data Mining Contest 2007 (winner in Task 1 and second-runner up in Task 2)
• 2007 Japanese Society for Artificial Intelligence Best Paper Award (JSAI Best Paper Award)
• 2004 SIG Research Award of Japanese Society for Artificial Intelligence (SIG Knowledge-based Systems)

Publications

Journal Papers

• Hisashi Kashima, Shoko Suzuki, Shohei Hido, Yuta Tsuboi, Toshihiro Takahashi, Tsuyoshi Ide, Rikiya Takahashi and Akira Tajima: A Semisupervised Approach Using Spatio-temporal Information for Indoor Location Estimation, In Qiang Yang, Sinno Jialin Pan and Vincent Wenchen Zheng, Estimating Location Using Wi-Fi, IEEE Intelligent Systems, vol. 23, no. 1, pp. 8-13, Jan/Feb, 2008.
• Shoko Suzuki, Yuta Tsuboi, Hisashi Kashima, Shohei Hido, Toshihiro Takahashi, Tsuyoshi Ide, Rikiya Takahashi and Akira Tajima: A Dimensionality Reduction Approach, In Qiang Yang, Sinno Jialin Pan and Vincent Wenchen Zheng, Estimating Location Using Wi-Fi, IEEE Intelligent Systems, vol. 23, no. 1, pp. 8-13, Jan/Feb, 2008.
• Hisashi Kashima: Risk-sensitive Learning via Minimization of Empirical Conditional Value-at-risk, IEICE Transaction on Information and Systems, Vol. E90-D, No. 12, pp. 2043-2052, 2007.
• Tetsuji Kuboyama, Hisashi Kashima, Kiyoko F. Aoki-Kinoshita, Kouichi Hirata, Hiroshi Yasuda: A Spectrum Tree Kernel, Journal of Japanese Society of Artificial Intelligence, Vol. 22, No. 2, 2007.
• Hisashi Kashima and Naoki Abe: A Parameterized Probabilistic Model of Network Evolution for Supervised Link Prediction, Journal of Japanese Society of Artificial Intelligence, Vol. 22, No. 2, 2007. [in Japanese]
• Hisashi Kashima, Tadashi Tsumura, Tsuyoshi Ide, Takahide Nogayama, Ryo Hirade, Hiroaki Etoh and Takeshi Fukuda: Network-Based Problem Detection for Distributed Systems, IEICE Transaction, Vol. J89-D, No. 2, 2006. [in Japanese]
• Hisashi Kashima, Hiroshi Sakamoto and Teruo Koyanagi: Tree Kernels, Journal of Japanese Society of Artificial Intelligence, Vol. 21, No. 1, 2006. [in Japanese]
• Tetsuo Shibuya, Hisashi Kashima and Akihiko Konagaya: Efficient Filtering Methods for Clustering cDNAs with Spliced Sequence Alignment, Bioinformatics, 2003.
• Takanori Fukao, Hisashi Kashima and Norihiko Adachi: Decentralized Adaptive Control with Improved Transient Performance, Transactions of the Society of Instrument and Control Engineers, Vol.35, No.7, pp. 869-878, 1999

Book Chapters/Tutorials

• Hisashi Kashima: Kernel Methods for Analyzing Structured Data, Technical Report of IEICE "Pattern Recognition and Media Understanding", 2005 [in Japanese]
• Hisashi Kashima: Kernel Methods for Mining Structured Data, IPSJ Magazine, Vol. 46, No. 1, 2005 [in Japanese]
• Hisashi Kashima, Koji Tsuda and Akihiro Inokuchi: Kernels for Graphs, Kernel Methods in Computational Biology, MIT Press, 2004

Conference Papers (Refereed)

• Shohei Hido, Tsuyoshi Ide, Hisashi Kashima, Harunobu Kubo and Hirofumi Matsuzawa: Unsupervised change analysis using supervised learning, In Proc. 12th Pacific-Asia Conference on Knowledge Discovery and Data Mining (PAKDD2008), 2008.
• Tsuyoshi Kato, Hisashi Kashima and Masashi Sugiyama: Integration of multiple networks for robust label propagation, In Proc. 2008 SIAM International Conference on Data Mining SDM2008), 2008.
• Shohei Hido and Hisashi Kashima: Roughly balanced bagging for imbalanced data, In Proc. 2008 SIAM International Conference on Data Mining (SDM2008), 2008.
• Yuta Tsuboi, Hisashi Kashima, Shohei Hido, Steffen Bickel and Masashi Sugiyama: Direct density ratio estimation for large-scale covariate shift adaptation, In Proc. 2008 SIAM International Conference on Data Mining (SDM2008), 2008.
• Hisashi Kashima, Shoko Suzuki, Shohei Hido, Yuta Tsuboi, Toshihiro Takahashi, Tsuyoshi Ide, Rikiya Takahashi and Akira Tajima: A Semi-supervised Approach to Indoor Location Estimation, In IEEE ICDM Data Mining Contest, Omaha NE, USA, 2007 (winner for Task 1 out of 15 teams)
• Shoko Suzuki, Yuta Tsuboi, Hisashi Kashima, Shohei Hido, Toshihiro Takahashi, Tsuyoshi Ide, Rikiya Takahashi and Akira Tajima: A Semi-supervised Approach to Transferring the Learned Knowledge for Indoor Location Estimation, In IEEE ICDM Data Mining Contest, Omaha NE, USA, 2007 (second runner up for Task 2 out of 17 teams)
• Masashi Sugiyama, Shinichi Nakajima, Hisashi Kashima, Paul von Bunau and Motoaki Kawanabe: Direct Importance Estimation with Model Selection and Its Application to Covariate Shift Adaptation, In Proc. 21st Annual Conference on Neural Information Processing Systems (NIPS2007), Vancouver, B.C., Canada, 2007.
• Tsuyoshi Kato, Hisashi Kashima, Masashi Sugiyama and Kiyoshi Asai: Multi-task Learning via Conic Programming, In Proc. 21st Annual Conference on Neural Information Processing Systems (NIPS2007), Vancouver, B.C., Canada, 2007.
• Tsuyoshi Ide and Hisashi Kashima: Effective Dimension in Anomaly Detection: Its Application to Computer Systems, New Frontiers in Artificial Intelligence (Post-proceedings of the Eighteenth Annual Conference of Japanese Society of Artificial Intelligence), Lecture Notes in Artificial Intelligence, Vol. 3609, pp.189-204, 2007.
• Tetsuji Kuboyama, Kouichi Hirata, Kiyoko F. Aoki-Kinoshita, Hisashi Kashima and Hiroshi Yasuda: A Gram Distribution Kernel Applied to Glycan Classification and Motif Extraction, In Proc. 17th International Conference on Genome Informatics (GIW2006), Yokohama, Japan, 2006.
• Hisashi Kashima and Naoki Abe: A Parameterized Probabilistic Model of Network Evolution for Supervised Link Prediction, In Proc. 6th IEEE International Conference on Data Mining (ICDM2006), Hong Kong, 2006.
• Tetsuji Kuboyama, Hisashi Kashima, Kiyoko F. Aoki-Kinoshita, Koichi Hirata and Hiroshi Yasuda: A Spectrum Tree Kernel, In Proc. The International Workshop on Data-Mining and Statistical Science (DMSS2006), Sapporo, Japan, 2006.
• Tetsuji Kuboyama, Kilho Shin and Hisashi Kashima: Flexible Tree Kernels Based on Counting the Number of Tree Mappings, In Proc. Workshop on Mining and Learning (held with ECML/PKDD 2006), Berlin, Germany, 2006.
• Hisashi Kashima: Risk-Sensitive Learning via Expected Shortfall Minimization, To apper in Proc. 2006 SIAM Conference on Data Mining (SDM06), Bethesda, Maryland, USA, 2006.
• Hisashi Kashima, Tadashi Tsumura, Tsuyosih Ide, Takahide Nogayama, Ryo Hirade, Hiroaki Etoh and Takeshi Fukuda: Network-Based Problem Detection for Distributed Systems, In Proc. 21st International Conference on Data Engineering (ICDE2005), Tokyo, Japan, 2005
• Tsuyoshi Ide and Hisashi Kashima: Eigenspace-based Anomaly Detection in Computer Systems, In Proc. 10th ACM SIGKDD Conference (KDD2004), Seattle, WA, USA, 2004
• Hisasih Kashima and Yuta Tsuboi: Kernel-Based Discriminative Learning Algorithms for Labeling Sequences, Trees and Graphs, In Proc. 21st International Conference on Machine Learning (ICML2004), Banff, Alberta, Canada, 2004
• Makoto Kano, Hisashi Kashima,Tetsuo Shibuya, Kaori Ide, Aiko Kashihara, Noriko Nakagawa, Mariko Hatakeyama, Seiki Kuramitsu and Akihiko Konagaya: A Method for Normalization of Gene Expression Data, In Proc. Genome Informatics Workshop (GIW2003), Yokohama, Japan, 2003
• A. Inokuchi and H. Kashima: Mining Significant Pairs of Patterns from Graph Structures with Class Labels, In Proc. 3rd IEEE International Conference on Data Mining (ICDM2003), Melbourne, Florida, USA, 2003
• Hisashi Kashima , Koji Tsuda and Akihiro Inokuchi: Marginalized Kernels Between Labeled Graphs, In Proc. 20th International Conference on Machine Learning (ICML2003), Washington, DC USA, 2003
• Hisashi Kashima and Akihiro Inokuchi: Kernels for Graph Classification, In Proc. 1st ICDM Workshop on Active Mining (AM-2002), Maebashi, Japan, 2002
• Hisashi Kashima and Teruo Koyanagi: Kernels for Semi-Structured Data, In Proc. 19th International Conference on Machine Learning (ICML2002), Sydney, Australia, 2002
• Takanori Fukao, Hisashi Kashima and Norihiko Adachi: Decentralized Adaptive Control of Dynamic Interconnected Systems with Improved Performance, In Proc. 8th IFAC Symposium on Large Scale Systems: Theory and Applications, pp. 138-143, 1998
• Takanori Fukao, Hisashi Kashima and Norihiko Adachi: Robust Adaptive Control of Large-Scale Systems with Unmodeled Dynamic Interconnections, In Proc. 2nd Asian Control Conference, Vol 2, pp. 5-8, 1997

Conference Papers (Unrefereed)

• Hisashi Kashima, Kazutaka Yamasaki, Hiroto Saigo and Akihiro Inokuchi: Regression with Intervals, The International Workshop on Data-Mining and Statistical Science (DMSS2007), 2007. [in Japanese]
• Masashi Sugiyama, Shinichi Nakajima, Hisashi Kashima, Paul von Bunau and Motoaki Kawanabe: Kullback-Leibler importance estimation procedure for covariate shift adaptation, The International Workshop on Data-Mining and Statistical Science (DMSS2007), 2007.
• Tetsuji Kuboyama, Koichi Hirata, Hisashi Kashima and Kiyoko F. Aoki-Kinoshita: The Gram Distribution Kernel: A Tree Kernel for Glycan Classification, In Technical Report of Japanese Society of Artificial Intelligence, SIG-FPAI, 2006.
• Hisashi Kashima: Risk-Sensitive Classification Learning, In Proc. 8th Information-Based Induction Sciences (IBIS 2005), 2005. [in Japanese]
• Yuta Tsuboi, Hisashi Kashima: Design of Discriminative Models for Labeling Structured Data, In Proc. 8th Information-Based Induction Sciences (IBIS 2005), 2005. [in Japanese]
• Tsuyoshi Ide and Hisashi Kashima: Eigenspace Approach to Anomaly Detection of Computer Systems, In Proc. Annual Conference of Japanese Society of Artificial Intelligence, 2004. [in Japanese]
• Hisashi Kashima and Yuta Tsuboi: Kernel-based Discriminative Learning Algorithms for Labeling Sequences, Trees and Graphs, In Technical Report of The Institute for Electronics, Information and Communication Engineers, AI, 2004. [in Japanese]
• Akihiro Inokuchi and Hisashi Kashima: Discovering Significant Pairs of Patterns from Graph Structures with Class Labels, In Technical Report of Japanese Society of Artificial Intelligence, SIG-KBS, 2004. [in Japanese]
• Tetsuo Shibuya, Chris Schoenbach, Hisashi Kashima and Akihiko Konagaya: Accurate cDNA Clustering Algorithm based on Spliced Sequence Alignment, In Technical Report of The Institute for Electronics, Information and Communication Engineers, COMP-2002-9-14, 17-24, 2002.
• Hisashi Kashima and Teruo Koyanagi: Kernels for Semi-Structured Data, In Technical Report of Japanese Society of Artificial Intelligence, SIG-FAI-A104, 2002. [in Japanese]
• Hisashi Kashima: Inferring Genetic Networks from Gene Expression Data Using Probabilistic Boolean Network Models, In Technical Report of Japanese Society of Artificial Intelligence, SIG-MBI, 2001. [in Japanese]
• Hisashi Kashima: Attribute Selection Using Multplicative Update, In Proc. Annual Conference of Japanese Society of Artificial Intelligence, 2001. [in Japanese]
• Hisashi Kashima and Yasumasa Kajinaga: Optimal Winner Determination Algorithms for E-procurement Auction, In Techinical Report of The Institute for Electronics, Information and Communication Engineers, COMP-2000-57-63, 17-23, 2000.

Technical Reports

• Hisashi Kashima and Yasumasa Kajinaga: Optimal Winner Determination Algorithms for E-procurement Auction, IBM Research Report, 2000
• Hisashi Kashima: Inferring Genetic Networks from Gene Expression Data Using Probabilistic Boolean Network Models, IBM Research Report, 2002
• Hisashi Kashima: On-line Kernel Principal Component Analysis, IBM Research Report, 2002
• Hisashi Kashima and Teruo Koyanagi: Kernels for Semi-structured Data, IBM Research Report, 2003
• Hisashi Kashima, Koji Tsuda and Akihiro Inokuchi: Marginalized Kernels Between Labeled Graphs, IBM Research Report, 2003