Academy of Mathematics and Systems Science, CAS Colloquia & Seminars | Speaker: | Dr. Yuto Miyatake, Osaka University | Inviter: | | Title: | Quantifying discretisation errors by isotonic regression and its application to estimating ODE models | Time & Venue: | 2019.09.09 11:35-12:20 N204 | Abstract: | One conventional approach of estimating ODE models from noisy observations is to fit numerical solutions of ODEs to data. However, such a method does not account for the discretisation error in numerical solutions and has limited estimation accuracy. In this talk, we present an estimation method with discretisation error quantification. Note that it is hoped that the discretisation error is quantified as efficiently as possible and when a structure-preserving method is employed its structure is effectively used. These are achieved by modelling the discretisation error as random variables and estimating their variances by an isotonic regression algorithm. This talk is based on joint work with Takeru Matsuda. | | | |