ML Anthology

Discovering Dynamics

Saso Dzeroski, Ljupco Todorovski
ICML 1993 pp. 97-103
doi:10.1016/B978-1-55860-307-3.50019-8 /icml/1993/dzeroski1993icml-discovering/

Abstract

Machine discovery is concerned with the task of finding laws from experimental and/or observational data. Existing machine discovery systems have mostly generated laws describing static situations. The paper presents LAGRANGE, a system that constructs a set of differential and/or algebraic equations that describe an observed behavior of a dynamic system. As such, LAGRANGE extends the scope of machine discovery to dynamic systems. We show that LAGRANGE is able to generate appropriate sets of laws for several nonlinear dynamic systems from traces of their behavior.

ICML Semantic Scholar

Cite

Text

Dzeroski and Todorovski. "Discovering Dynamics." International Conference on Machine Learning, 1993. doi:10.1016/B978-1-55860-307-3.50019-8

Markdown

[Dzeroski and Todorovski. "Discovering Dynamics." International Conference on Machine Learning, 1993.](https://mlanthology.org/icml/1993/dzeroski1993icml-discovering/) doi:10.1016/B978-1-55860-307-3.50019-8

BibTeX

@inproceedings{dzeroski1993icml-discovering,
  title     = {{Discovering Dynamics}},
  author    = {Dzeroski, Saso and Todorovski, Ljupco},
  booktitle = {International Conference on Machine Learning},
  year      = {1993},
  pages     = {97-103},
  doi       = {10.1016/B978-1-55860-307-3.50019-8},
  url       = {https://mlanthology.org/icml/1993/dzeroski1993icml-discovering/}
}