We aim to create new principles of machine learning and innovative technologies for the creation of innovative artificial intelligence services based on the utilization of big data generated in various scientific fields, industry, and the real world. Furthermore, through joint research, we will apply machine learning technology to the natural and social sciences, and contribute to the further advancement of machine learning technology and the development of scientific research.
Two of our papers, "Baxter permutation process" and "Meta-learning from Tasks with Heterogeneous Attribute Spaces," were accepted by NeurIPS2020.
Our paper, "Time-delayed collective flow diffusion models for inferring latent people flow from aggregated data at limited locations," was accepted by Artificial Intelligence.
Our paper, "Gamifying World Wide Web Using Web Browsers," was selected Spacially Selected Paper by Information Processing Society of Japan.
Our paper, "Reinforcement Learning in Latent Action Sequence Space," was accepted by IROS2020.