Publication of the cross stream ecosystem comparison.

Title:

Resource Type:

document --> guidance / decision support

Country:

EU Projects

Year:

2003

Author 1/Producer:

Paequann Consortium

Author / Producer Type:

EC Project

Report / download web link (=direct link):

http://aquaeco.ups-tlse.fr/index2.htm

Format (e.g. PDF):

PDF

EUGRIS Keyword(s):

Water resources and their management -->Water resources and their management Overview

Short description:

The PAEQANN project aims to review the current ecological theories that could help identify models for the prediction of the community structure in aquatic ecosystems. The report compares the performance of a number of classical statistical models, including artificial neural networks, support vector machines and dynamic models. Each model has different properties, with its own advantages and disadvantages. The exploration of community – environment relationships is strengthened when both multivariate and artificial neural network techniques were used. The prediction support vector machines were also promising. Further improvement could be achieved by adding a priori, well established knowledge into the analysis and prediction.

Link to Project(s):

PAEQANN Predicting Aquatic Ecosystem Quality using Artificial Neural Networks: Impact of Environmental characteristics on the Structure of Aquatic Communities (Algae, Benthic and Fish Fauna).

Submitted By:

Dr Stefan Gödeke WhoDoesWhat? Last update: 19/07/2007