Logo: to the web site of Uppsala University

uu.sePublications from Uppsala University
EnglishSvenskaNorsk
Change search
ExportLink to record
Permanent link

Direct link
BETA

Project

Project type/Form of grant
Project grant
Title [sv]
Inlärning av flexibla modeller av olinjär dynamik
Title [en]
Learning flexible models of nonlinear dynamics
Abstract [en]
Learning systems of the kind we develop in this project constitute core technology of all model-based engineered systems acting and interacting autonomously in a physical environment. This project will result in the development of improved probabilistic models of complex nonlinear dynamic phenomena, together with the algorithms allowing us to learn these models from measured data. This is a capability that is fundamental to many scientific fields and its importance is constantly growing due to the large quantities of data that are now becoming available.The research area within which we are active is indeed vast, but our research program is focused. It consists in three strategically linked subprojects: 1) Developing flexible probabilistic models capable of enforcing fundamental dynamical behaviors; 2) Derivation of new nonlinear stochastic optimizers that use curvature information by exploiting non-parametric models; 3) Developing coupling constructions of computational learning algorithms, allowing them to interact and in that way become more suitable to use as components within larger algorithms.This project marks a new research direction compared with our ongoing VR project. Importantly, we have established the necessary collaborations allowing us to pursue this project together with leading researchers around the world.
Principal InvestigatorSchön, Thomas
Coordinating organisation
Uppsala University
Funder
Period
2018-01-01 － 2021-12-31
National Category
Control Engineering
Identifiers
DiVA, id: project:6002Project, id: 2017-03807_VR

Search in DiVA

Control Engineering

Search outside of DiVA

GoogleGoogle Scholar
v. 2.47.0
|
WCAG
|
Uppsala University Library
|
Ask the Library
|
Log in to DiVA
|
Search and link in DiVA