The Predictive Medicine Special Project develops methods to interpret and predict life phenomena that have been difficult to understand in the past by incorporating AI and data science technologies, which have been growing rapidly in recent years.
Humans, nature, and society share the common characteristic of being nonlinear, non-equilibrium, multi-body, open systems. To date, the natural sciences have attempted to simulate such realities by simulating the laws that govern events. In contrast, surrogate models derived from real-world data show significantly higher predictive accuracy than traditional methods, even though they do not incorporate the laws governing events. However, surrogate models cannot adequately explain "why" such changes occur. By developing a new inference framework that overcomes these challenges, we aim to realize healthcare that prevents the onset of disease.
Kazuhiro Sakurada
Project Director, Predictive Medicine Special Project, RIKEN Center for Integrative Medical Sciences