Guest Editors’ Introduction: special issue of selected papers from ECML PKDD 2011

Gunopulos, Dimitrios; Malerba, Donato; Vazirgiannis, Michalis

doi:10.1007/s10618-012-0282-x

Cited by 1 publication

(1 citation statement)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Our results (Figures 2, 3, 4) are the outputs of a complex model that includes a linear environmental structure with eight environmental covariates and a deep neural network spatial structure with 30 x 2 layers. Complex models run a risk of overfitting, so to estimate the risk of overfitting after elastic-net regularisation, we performed a 20-fold cross-validation test using stratified multi-label sampling to partition the data (Gunopulos et al . 2011; Szymański & Kajdanowicz 2017).…”

Section: Methodsmentioning

confidence: 99%

Assembly processes inferred from eDNA surveys of a pond metacommunity are consistent with known species ecologies

Cai,

Pichler,

Biggs

et al. 2023

Preprint

View full text Add to dashboard Cite

Metacommunity theory unites community and spatial ecology. Recent innovations in sequencing technology may now allow us, for the first time, to confront this theory with entire trophic community data at large scales. We metabarcoded vertebrate eDNA from 320 ponds in England. Using this ‘novel community data’, we calculated the ‘internal structure’ of this metacommunity, which is the contributions of environmental, biotic, and spatial processes, as well as stochasticity, to the distribution of each species and the composition of each site. Our results suggest that community assembly processes in this pondscape do not conform to a single metacommunity paradigm, but rather vary among species and over space. Combining this variation with external knowledge of species biologies allows us to better understand the differential influences of environmental filtering, species interactions, and dispersal limitation on several of the vertebrate species, paving the way for the routine empirical study of metacommunities.

show abstract

Section: Methodsmentioning

confidence: 99%