Ichigaku Takigawa
Ichigaku Takigawa connects machine learning with scientific discovery. His research focuses on learning with discrete structures and discovery in data-rich natural sciences. He directs RIKEN’s Chemical Reaction Informatics Team. His shared research includes frequent-subgraph mining, sparse learning with graph features, catalyst analysis and machine-learning methods for microscopy. He coauthored work on extrapolative learning for catalyst discovery in 2023. His research uses chemistry as a setting for questions about representation, generalization and learning from limited observations.
Timeline
-
2011–2017
Learning with graph structure
Takigawa and Hiroshi Mamitsuka coauthored papers on frequent-subgraph mining in 2011 and sparse learning over a complete subgraph feature set in 2017. These connect discrete structures with statistical models.
-
2018–2020
Learning for catalysis
Takigawa coauthored work on predicting adsorption energies on metal alloys, analyzing heterogeneous catalysts from published data and reviewing machine learning for catalysis informatics.
-
2022–2023
Images and catalyst discovery
Takigawa coauthored papers on detecting nucleation and improving low-dose microscopy images through learning. A 2023 paper studied extrapolative learning for discovery of multi-element catalysts.
-
As of October 2026
Chemical Reaction Informatics Team
Takigawa is team director; the team studies learning for discovery and design of chemical reactions. His research focuses on discrete structures and machine discovery.
Sources and credits
Photo: www.riken.jp; creator not stated. License: Not stated by source. Source: riken.jp. Cropped to a square and resized. Ask us to remove or replace this photo.