报告摘要：

Nowadays, multi-disciplinary integrated assessment models are widely used in studies of complex socio-encironmental systems. In an integrated assessment model, each disciplinary submodel is often represented as an ordinary differential equation designed and assessed by an expert in the corresponding discipline. Typically, each disciplinary submodel is equipped with a parameter.

Fitting the entire integrated assessment model is understood as finding the parameter values for the submodels, which provide the best fit of the entire model's trajectory with historical data. The best-fit parameter values are used to form a forecast by extending the model's trajectory into the future. As usual, for every disciplinary submodel, the paramter values found through fitting the entire model do not fit the disciplinary data best, which makes the disciplinary experts view the fitting result as a non-reliable one. The lecture introduces a notion of equilibrium-fit disciplinary parameter values establishing a consensus among the diciplinary experts, and argues briefly on the existence and construction of equilibrium-fit parameter values.

 

报告人简介：

Arkady Kryazhimskiy博士是国际知名的应用数学家，俄罗斯科学院院士，是清华大学海外高水平合作学者支持项目聘请来访的海外知名学者。Kryazhimskiy博士目前担任俄罗斯科学院Steklov数学研究所资深研究员，同时兼任维也纳国际应用系统分析研究所(IIASA)先进系统分析研究部资深研究员、莫斯科大学教授。Kryazhimskiy博士1971年毕业于前苏联乌拉尔国立大学数学专业，1981年获得物理和数学正博士学位，1997年当选俄罗斯科学院通讯院士，2006年正式当选俄罗斯科学院院士。他曾多年担任维也纳国际应用系统分析研究所(IIASA)动力系统研究部和先进系统分析研究部的负责人。Kryazhimskiy博士的主要研究兴趣包括优化方法、最优控制、不适定和反问题、动态博弈论以及这些方法在经济、环境领域的应用。