Takuya Konishi

Specially-appointed lecturer

Graduate School of Information Science and Technology

The University of Osaka

Email: konishi[ a t ]ist.osaka-u.ac.jp

Research interests

• Machine learning
• Knowledge discovery and data mining

Publications

• Kernel Occupation Readout for Oscillatory Recurrent Neural Networks. Yuto Inui, Masahiro Ikeda, Takuya Konishi, Yoshinobu Kawahara. IJCNN 2026. to appear.
• Predictive Fair Representation Learning with Variational Autoencoders. Tatsuya Yamada, Takuya Konishi, Yoshinobu Kawahara. New Generation Computing 2026. [link]
• Explorative Curriculum Learning for Strongly Correlated Electron Systems. Kimihiro Yamazaki, Takuya Konishi, Yoshinobu Kawahara. ACM TAIS 2026. [link]
• Learning with Almost Invariant Sets in Neural Oscillatory ODEs. Yuto Inui, Takuya Konishi, Yoshinobu Kawahara. ICONIP 2024. [link]
• Stable Invariant Models via Koopman Spectra. Takuya Konishi, Yoshinobu Kawahara. Neural Networks 2023. [link]
• End-to-End Learning for Prediction and Optimization with Gradient Boosting. Takuya Konishi, Takuro Fukunaga. ECML-PKDD 2020. [link]
• Stochastic Submodular Maximization with Performance-Dependent Item Costs. Takuro Fukunaga, Takuya Konishi, Sumio Fujita, Ken-ichi Kawarabayashi. AAAI 2019. [link]
• Identifying Key Observers to Find Popular Information in Advance. Takuya Konishi, Tomoharu Iwata, Kohei Hayashi, Ken-ichi Kawarabayashi. IJCAI 2016. [link]
• Extracting Search Query Patterns via the Pairwise Coupled Topic Model. Takuya Konishi, Takuya Ohwa, Sumio Fujita, Kazushi Ikeda, Kohei Hayashi. WSDM 2016. [link]
• Variational Bayesian Inference Algorithms for Infinite Relational Model of Network Data. Takuya Konishi, Takatomi Kubo, Kazuho Watanabe, Kazushi Ikeda. IEEE TNNLS 2015. [link]
• Topic Model for User Reviews with Adaptive Windows. Takuya Konishi, Fuminori Kimura, Akira Maeda. ECIR 2013 (short paper). [link]
• Estimating Aspects in Online Reviews Using Topic Model with 2-Level Learning. Takuya Konishi, Taro Tezuka, Fuminori Kimura, Akira Maeda. IMECS 2012. [link]

Workshop presentations

• Weight-sharing Transformer quantum states with Suzuki–Trotter decompositions. Kimihiro Yamazaki, Itsushi Sakata, Takuya Konishi, Yoshinobu Kawahara. Machine Learning and the Physical Sciences Workshop at the 39th conference on Neural Information Processing Systems (NeurIPS), 2025. [link]
• Variational Bayesian Inference Algorithms for Network Infinite Relational Model. Takuya Konishi, Takatomi Kubo, Kazuho Watanabe, Kazushi Ikeda. NIPS 2013 Workshop on Frontiers of Network Analysis: Methods, Models, and Applications. [link]

Preprints

• Physics-inspired transformer quantum states via latent imaginary-time evolution. Kimihiro Yamazaki, Itsushi Sakata, Takuya Konishi, Yoshinobu Kawahara. arXiv:2602.03031. [link]
• A Tractable Fully Bayesian Method for the Stochastic Block Model. Kohei Hayashi, Takuya Konishi, Tatsuro Kawamoto. arXiv:1602.02256. [link]

Work experience

• April 2023 - current. Specially-appointed lecturer at Graduate School of Information Science and Technology, The University of Osaka.
• April 2021 - current. Visiting scientist at Center for Advanced Intelligence Project, RIKEN.
• April 2022 - March 2023. Assistant professor at Institute of Mathematics for Industry, Kyushu University.
• November 2020 - March 2022. Project assistant professor at Institute of Mathematics for Industry, Kyushu University.
• April 2015 - October 2020. Project researcher at Global Research Center for Big Data Mathematics, National Institute of Informatics.

Education

• April 2012 - March 2015. Ph.D. of Engineering from Graduate School of Information Science, Nara Institute of Science and Technology.
• April 2010 - March 2012. Master of Engineering from Information Science and Engineering, Graduate School of Science and Engineering, Ritsumeikan University.
• April 2006 - March 2010. Bachelor of Engineering from Department of Media Technology, College of Information Science and Engineering, Ritsumeikan University.

Award

• Best Student Paper Award. The 2012 IAENG International Conference on Data Mining and Applications.

Review activities

• AAAI, ACML, AISTATS, ICML, ICLR, IEEE TNNLS, IJCAI, NeurIPS, etc.