articlenews
NEWS
articlenews
【Article】Physics-informed active learning for the design of thermally conductive composite microstructures with controlled anisotropy
articlenews
【Article】Machine Learning-Based Prediction of Polymer Chemical Resistance to Organic Solvents
articlenews
【Article】Database and deep-learning scalability of anharmonic phonon properties by automated brute-force first-principles calculations
newspresentation
【Lecture】日本化学会 第106春季年会 (2026)
MISSION
Prediction and Discovery of Innovative Materials through Data Science
The material landscape is extremely vast. For example, in the chemical space of organic compounds, it is estimated that there exist more 1060 potential materials yet to be discovered. Our mission is to uncover innovative materials from this vast unexplored space, using advanced data science technologies as a key tool.
Creative Prediction and Design through Data Science
"Innovative materials" exist in a blue ocean where data does not yet exist. Therefore, it is hard to reach there solely through interpolative predictions based on existing data. Our aim is to achieve "extrapolative discovery”. To realize this, we systematize a closed-loop process that integrates high-throughput experiment, computer experiments, machine learning, and hypothesis generation, demonstrating our concepts through the discovery of new materials across various domains.
Establishing Open Data Platform through Industry-Academia Collaboration
In data-driven research, data is the most crucial resource. It is a power game where big data holders become winners. However, compared to other fields, materials research suffers from a significant shortage of data. To break through this limitation, we are forming a consortium involving numerous researchers from both academia and industry to develop the world's largest materials database.


