Institutional Repository of Vilnius University
Finding analytical approximations of differential equations using machine learning and symbolic computation /
Abstract
dc:descriptionThis work addresses the issue of finding approximations for differential equations. The mathematical models that use differential equations have various parameters which are changed depending on the actual experiment. A novel method of solving the modeled equations is presented, which takes the differentiable architecture search (DARTS) method and extends it with symbolic integration, where equations are solved for a particular interval of parameters. Unlike traditional numerical methods (e.g. Euler method, finite difference method), the model does not need to be rerun for each change in the parameters. Experiments show that the architecture is capable of approximating the Malthusian growth model equation and Michaelis–Menten kinetics equations for some particular parameter intervals.
Degree
thesis:*- Grantor dc:publisher
- Institutional Repository of Vilnius University
- Year dc:date
- 2022
Author and committee
dc:creator, dc:contributor.*- Author dc:creator
-
- Sasnauskas, Paulius,
Subjects
dc:subject × 1Rights
dc:rights- Statement dc:rights
-
- info:eu-repo/semantics/openAccess
- Language dc:language
- eng
Identifiers
dc:identifier.*- Repository record dc:identifier
- https://repository.vu.lt/VU:ELABAETD192980568&prefLang=en_US