Math
Institute of Statistical Mathematics Symposium 'AI4S 2026': Machine Learning and Statistical Mathematics Open Up 'AI for Science'
Related terms: 数学 統計 機械学習 AI 確率統計 不確実性
Key points (reviewing AI explanation)
Behind the AI boom lies a mass of unglamorous mathematics (probability, statistics, and linear algebra). Machine learning is the task of 'finding the function that best fits a large amount of data,' and how good that fit is depends on the mathematics of 'how you measure error.' What's especially important in AI for Science is expressing 'how certain something is (uncertainty)' as a number. In medical and materials research, honestly estimating the possibility of being wrong is more of a lifeline than simply being right. This story shows how mathematics serves as the 'guardian that makes AI trustworthy.'
Explanation
The Institute of Statistical Mathematics is planning a symposium titled 'AI4S 2026: New Developments in AI for Science Pioneered by Machine Learning and Statistical Mathematics.' AI for Science refers to efforts to accelerate research itself—such as drug discovery, materials development, and weather forecasting—using AI. The foundation for this lies in statistical mathematics and machine learning, which uncover laws from data. Mathematics that handles probability and error is indispensable for AI to produce not just 'plausible-sounding answers' but 'trustworthy answers.'
Source: www.ism.ac.jp