dorsal/arxiv
View SchemaStatistical Predictive Models in Ecology: Comparison of Performances and Assessment of Applicability
| Authors | Can Ozan Tan, Uygar Ozesmi, Meryem Beklioglu, Esra Per, Bahtiyar Kurt |
|---|---|
| Categories | |
| ArXiv ID | q-bio/0510031 |
| URL | https://arxiv.org/abs/q-bio/0510031 |
| DOI | 10.1016/j.ecoinf.2006.03.002 |
| Journal | Ecological Informatics, 1:195-211. 2006 |
Abstract
Ecological systems are governed by complex interactions which are mainly nonlinear. In order to capture this complexity and nonlinearity, statistical models recently gained popularity. However, although these models are commonly applied in ecology, there are no studies to date aiming to assess the applicability and performance. We provide an overview for nature of the wide range of the data sets and predictive variables, from both aquatic and terrestrial ecosystems with different scales of time-dependent dynamics, and the applicability and robustness of predictive modeling methods on such data sets by comparing different statistical modeling approaches. The methods considered k-NN, LDA, QDA, generalized linear models (GLM) feedforward multilayer backpropagation networks and pseudo-supervised network ARTMAP. For ecosystems involving time-dependent dynamics and periodicities whose frequency are possibly less than the time scale of the data considered, GLM and connectionist neural network models appear to be most suitable and robust, provided that a predictive variable reflecting these time-dependent dynamics included in the model either implicitly or explicitly. For spatial data, which does not include any time-dependence comparable to the time scale covered by the data, on the other hand, neighborhood based methods such as k-NN and ARTMAP proved to be more robust than other methods considered in this study. In addition, for predictive modeling purposes, first a suitable, computationally inexpensive method should be applied to the problem at hand a good predictive performance of which would render the computational cost and efforts associated with complex variants unnecessary.
{
"annotation_id": "8a384b59-340c-4b21-b282-53489ceb7b00",
"date_created": "2026-03-02T18:01:32.331000Z",
"date_modified": "2026-03-02T18:01:32.331000Z",
"file_hash": "41d2095a3fb796d004175a1e4ac912ba3e46db32c784705a6b26ea26bbec129a",
"private": false,
"record": {
"abstract": "Ecological systems are governed by complex interactions which are mainly\nnonlinear. In order to capture this complexity and nonlinearity, statistical\nmodels recently gained popularity. However, although these models are commonly\napplied in ecology, there are no studies to date aiming to assess the\napplicability and performance. We provide an overview for nature of the wide\nrange of the data sets and predictive variables, from both aquatic and\nterrestrial ecosystems with different scales of time-dependent dynamics, and\nthe applicability and robustness of predictive modeling methods on such data\nsets by comparing different statistical modeling approaches. The methods\nconsidered k-NN, LDA, QDA, generalized linear models (GLM) feedforward\nmultilayer backpropagation networks and pseudo-supervised network ARTMAP. For\necosystems involving time-dependent dynamics and periodicities whose frequency\nare possibly less than the time scale of the data considered, GLM and\nconnectionist neural network models appear to be most suitable and robust,\nprovided that a predictive variable reflecting these time-dependent dynamics\nincluded in the model either implicitly or explicitly. For spatial data, which\ndoes not include any time-dependence comparable to the time scale covered by\nthe data, on the other hand, neighborhood based methods such as k-NN and ARTMAP\nproved to be more robust than other methods considered in this study. In\naddition, for predictive modeling purposes, first a suitable, computationally\ninexpensive method should be applied to the problem at hand a good predictive\nperformance of which would render the computational cost and efforts associated\nwith complex variants unnecessary.",
"arxiv_id": "q-bio/0510031",
"authors": [
"Can Ozan Tan",
"Uygar Ozesmi",
"Meryem Beklioglu",
"Esra Per",
"Bahtiyar Kurt"
],
"categories": [
"q-bio.QM"
],
"doi": "10.1016/j.ecoinf.2006.03.002",
"journal_ref": "Ecological Informatics, 1:195-211. 2006",
"title": "Statistical Predictive Models in Ecology: Comparison of Performances and Assessment of Applicability",
"url": "https://arxiv.org/abs/q-bio/0510031"
},
"schema_id": "dorsal/arxiv",
"source": {
"execution_id": "da050bcf-93f1-46f7-bafc-53fe24077f99",
"id": "arXiv Dataset IDs",
"type": "Model",
"variant": "snapshot-2026-03-01",
"version": "0.1.0"
},
"user_id": 1000002
}