Fast fitting of neural ordinary differential equations by Bayesian neural gradient matching to infer ecological interactions from time‐series data
Abstract Inferring ecological interactions is hard because we often lack suitable parametric representations to portray them. Neural ordinary differential equations (NODEs) provide a way of estimating interactions non‐parametrically from time‐series data. NODEs, however, are slow to fit, and inferre...
Published in: | Methods in Ecology and Evolution |
---|---|
Main Authors: | , |
Format: | Article in Journal/Newspaper |
Language: | English |
Published: |
Wiley
2023
|
Subjects: | |
Online Access: | https://doi.org/10.1111/2041-210X.14121 https://doaj.org/article/3279e3b73e7f4fab874959a43966a6a2 |
id |
ftdoajarticles:oai:doaj.org/article:3279e3b73e7f4fab874959a43966a6a2 |
---|---|
record_format |
openpolar |
spelling |
ftdoajarticles:oai:doaj.org/article:3279e3b73e7f4fab874959a43966a6a2 2023-08-27T04:12:35+02:00 Fast fitting of neural ordinary differential equations by Bayesian neural gradient matching to infer ecological interactions from time‐series data Willem Bonnaffé Tim Coulson 2023-06-01T00:00:00Z https://doi.org/10.1111/2041-210X.14121 https://doaj.org/article/3279e3b73e7f4fab874959a43966a6a2 EN eng Wiley https://doi.org/10.1111/2041-210X.14121 https://doaj.org/toc/2041-210X 2041-210X doi:10.1111/2041-210X.14121 https://doaj.org/article/3279e3b73e7f4fab874959a43966a6a2 Methods in Ecology and Evolution, Vol 14, Iss 6, Pp 1543-1563 (2023) artificial neural networks ecological dynamics ecological interactions Geber method gradient matching microcosm Ecology QH540-549.5 Evolution QH359-425 article 2023 ftdoajarticles https://doi.org/10.1111/2041-210X.14121 2023-08-06T00:47:09Z Abstract Inferring ecological interactions is hard because we often lack suitable parametric representations to portray them. Neural ordinary differential equations (NODEs) provide a way of estimating interactions non‐parametrically from time‐series data. NODEs, however, are slow to fit, and inferred interactions usually are not compared with the ground truth. We provide a fast NODE fitting method, Bayesian neural gradient matching (BNGM), which relies on interpolating time series with neural networks and fitting NODEs to the interpolated dynamics with Bayesian regularisation. We test the accuracy of the approach by inferring ecological interactions in time series generated by an ODE model with known interactions. We compare these results against three existing approaches for estimating ecological interactions, standard NODEs, ODE models and convergent cross‐mapping (CCM). We also infer interactions in experimentally replicated time series of a microcosm featuring an algae, flagellate and rotifer population, in the hare and lynx system, and the Maizuru Bay community featuring 11 species. Our BNGM approach allows us to reduce the fitting time of NODE systems to only a few seconds and provides accurate estimates of ecological interactions in the artificial system, as true ecological interactions are recovered with minimal error. Our benchmark analysis reveals that our approach is both faster and more accurate than standard NODEs and parametric ODEs, while CCM was found to be faster but less accurate. The analysis of the replicated time series reveals that only the strongest interactions are consistent across replicates, while the analysis of the Maizuru community shows the strong negative impact of the chameleon goby on most species of the community, and a potential indirect negative effect of temperature by favouring goby population growth. Overall, NODEs alleviate the need for a mechanistic understanding of interactions, and BNGM alleviates the heavy computational cost. This is a crucial step availing quick NODE ... Article in Journal/Newspaper Lynx Rotifer Directory of Open Access Journals: DOAJ Articles Methods in Ecology and Evolution 14 6 1543 1563 |
institution |
Open Polar |
collection |
Directory of Open Access Journals: DOAJ Articles |
op_collection_id |
ftdoajarticles |
language |
English |
topic |
artificial neural networks ecological dynamics ecological interactions Geber method gradient matching microcosm Ecology QH540-549.5 Evolution QH359-425 |
spellingShingle |
artificial neural networks ecological dynamics ecological interactions Geber method gradient matching microcosm Ecology QH540-549.5 Evolution QH359-425 Willem Bonnaffé Tim Coulson Fast fitting of neural ordinary differential equations by Bayesian neural gradient matching to infer ecological interactions from time‐series data |
topic_facet |
artificial neural networks ecological dynamics ecological interactions Geber method gradient matching microcosm Ecology QH540-549.5 Evolution QH359-425 |
description |
Abstract Inferring ecological interactions is hard because we often lack suitable parametric representations to portray them. Neural ordinary differential equations (NODEs) provide a way of estimating interactions non‐parametrically from time‐series data. NODEs, however, are slow to fit, and inferred interactions usually are not compared with the ground truth. We provide a fast NODE fitting method, Bayesian neural gradient matching (BNGM), which relies on interpolating time series with neural networks and fitting NODEs to the interpolated dynamics with Bayesian regularisation. We test the accuracy of the approach by inferring ecological interactions in time series generated by an ODE model with known interactions. We compare these results against three existing approaches for estimating ecological interactions, standard NODEs, ODE models and convergent cross‐mapping (CCM). We also infer interactions in experimentally replicated time series of a microcosm featuring an algae, flagellate and rotifer population, in the hare and lynx system, and the Maizuru Bay community featuring 11 species. Our BNGM approach allows us to reduce the fitting time of NODE systems to only a few seconds and provides accurate estimates of ecological interactions in the artificial system, as true ecological interactions are recovered with minimal error. Our benchmark analysis reveals that our approach is both faster and more accurate than standard NODEs and parametric ODEs, while CCM was found to be faster but less accurate. The analysis of the replicated time series reveals that only the strongest interactions are consistent across replicates, while the analysis of the Maizuru community shows the strong negative impact of the chameleon goby on most species of the community, and a potential indirect negative effect of temperature by favouring goby population growth. Overall, NODEs alleviate the need for a mechanistic understanding of interactions, and BNGM alleviates the heavy computational cost. This is a crucial step availing quick NODE ... |
format |
Article in Journal/Newspaper |
author |
Willem Bonnaffé Tim Coulson |
author_facet |
Willem Bonnaffé Tim Coulson |
author_sort |
Willem Bonnaffé |
title |
Fast fitting of neural ordinary differential equations by Bayesian neural gradient matching to infer ecological interactions from time‐series data |
title_short |
Fast fitting of neural ordinary differential equations by Bayesian neural gradient matching to infer ecological interactions from time‐series data |
title_full |
Fast fitting of neural ordinary differential equations by Bayesian neural gradient matching to infer ecological interactions from time‐series data |
title_fullStr |
Fast fitting of neural ordinary differential equations by Bayesian neural gradient matching to infer ecological interactions from time‐series data |
title_full_unstemmed |
Fast fitting of neural ordinary differential equations by Bayesian neural gradient matching to infer ecological interactions from time‐series data |
title_sort |
fast fitting of neural ordinary differential equations by bayesian neural gradient matching to infer ecological interactions from time‐series data |
publisher |
Wiley |
publishDate |
2023 |
url |
https://doi.org/10.1111/2041-210X.14121 https://doaj.org/article/3279e3b73e7f4fab874959a43966a6a2 |
genre |
Lynx Rotifer |
genre_facet |
Lynx Rotifer |
op_source |
Methods in Ecology and Evolution, Vol 14, Iss 6, Pp 1543-1563 (2023) |
op_relation |
https://doi.org/10.1111/2041-210X.14121 https://doaj.org/toc/2041-210X 2041-210X doi:10.1111/2041-210X.14121 https://doaj.org/article/3279e3b73e7f4fab874959a43966a6a2 |
op_doi |
https://doi.org/10.1111/2041-210X.14121 |
container_title |
Methods in Ecology and Evolution |
container_volume |
14 |
container_issue |
6 |
container_start_page |
1543 |
op_container_end_page |
1563 |
_version_ |
1775356894910611456 |