We gratefully acknowledge support from
the Simons Foundation and member institutions.
Full-text links:

Download:

Current browse context:

q-bio.QM

Change to browse by:

References & Citations

Bookmark

(what is this?)
CiteULike logo BibSonomy logo Mendeley logo del.icio.us logo Digg logo Reddit logo ScienceWISE logo

Quantitative Biology > Quantitative Methods

Title: Bayesian uncertainty quantification for data-driven equation learning

Abstract: Equation learning aims to infer differential equation models from data. While a number of studies have shown that differential equation models can be successfully identified when the data are sufficiently detailed and corrupted with relatively small amounts of noise, the relationship between observation noise and uncertainty in the learned differential equation models remains unexplored. We demonstrate that for noisy data sets there exists great variation in both the structure of the learned differential equation models as well as the parameter values. We explore how to combine data sets to quantify uncertainty in the learned models, and at the same time draw mechanistic conclusions about the target differential equations. We generate noisy data using a stochastic agent-based model and combine equation learning methods with approximate Bayesian computation (ABC) to show that the correct differential equation model can be successfully learned from data, while a quantification of uncertainty is given by a posterior distribution in parameter space.
Subjects: Quantitative Methods (q-bio.QM)
Cite as: arXiv:2102.11629 [q-bio.QM]
  (or arXiv:2102.11629v4 [q-bio.QM] for this version)

Submission history

From: Simon Martina-Perez [view email]
[v1] Tue, 23 Feb 2021 11:08:30 GMT (20090kb,D)
[v2] Mon, 17 May 2021 16:06:44 GMT (47105kb,D)
[v3] Mon, 7 Jun 2021 17:19:44 GMT (12776kb,D)
[v4] Wed, 29 Sep 2021 16:46:41 GMT (13084kb,D)

Link back to: arXiv, form interface, contact.