学术信息

Theory + Data: Building Hybrid Models with Stoichiometry and Machine Learning

主讲人简介:

王皓,加拿大生物数学首席科学家(Tier 1 Canada Research Chair in Mathematical Biosciences),加拿大阿尔伯塔大学终身正教授,阿尔伯塔大学数学生态学和传染病学交叉研究实验室主任。现任9个国际生物数学和动力系统主流杂志的主编或编委,已在高水平SCI期刊发表论文190余篇。在化学计量学,动物认知移动,环境污染毒素,微生物分解,物种入侵,传染病传播机制和预测等都做出了开创性和突破性研究。目前主导加拿大Alliance Missions等多个国家重点和重大基金。荣获加拿大应用与工业数学会科研突破奖(CAIMS-SCMAI Research Prize),阿尔伯塔大学杰出导师奖,约瑟夫·米歇尔指导奖,美国数学生物科学研究所杰出青年学者奖,以及加拿大国家基金科研加速资助奖等。

内容摘要:

Stoichiometric principles provide a rigorous, law-based backbone for building mechanistic models that are robust and empirically testable through conservation and physiological constraints. In this talk, I will introduce stoichiometric models that resolve key biological paradoxes by explicitly tracking elemental and quota-driven limitations, then demonstrate how data-driven components can (i) learn missing processes, (ii) improve short-term forecasting, and (iii) enable real-time monitoring. Methane biogenesis in oil sands tailings serves as a central application: we couple stoichiometric biodegradation dynamics with ML-enabled monitoring to better predict emissions trajectories. This research contributes to the broader goal of carbon neutrality.


关于东华大学

东华大学是教育部直属、国家“211工程”、国家“双一流”建设高校。学校秉承“崇德博学、砺志尚实”的校训,不断开拓奋进,
已发展成为以纺织、材料、设计为优势,特色鲜明的多科性、高水平大学。

了解更多