Predicting the Distribution of Fungal Crop Diseases from Abiotic and Biotic Factors Using Multi-Layer Perceptrons

Autor:	Michael J. Watts, Susan P. Worner
Rok vydání:	2009
Předmět:	Abiotic component Biotic component Host (biology) business.industry Ecology fungi food and beverages Climatic variables Distribution (economics) Biology Perceptron Crop business Multi layer
Zdroj:	Advances in Neuro-Information Processing ISBN: 9783642024894 ICONIP (1)
DOI:	10.1007/978-3-642-02490-0_110
Popis:	Predictions of the distribution of fungal crop diseases have previously been made solely from climatic data. To our knowledge there has been no study that has used biotic variables, either alone or in combination with climate factors, to make broad scale predictions of the presence or absence of fungal species in particular regions. The work presented in this paper used multi-layer perceptrons (MLP) to predict the presence and absence of several species of fungal crop diseases across world-wide geographical regions. These predictions were made using three sets of variables: abiotic climate variables; biotic variables, represented by host plant assemblages; And finally the combination of predictions of the climate and host assemblage MLP using a cascaded MLP architecture, such that final predictions were made from both abiotic and biotic factors.
Databáze:	OpenAIRE
Externí odkaz:	https://explore.openaire.eu/search/publication?articleId=doi_________::04152179e4a187203579d44c5edf4a99 https://doi.org/10.1007/978-3-642-02490-0_110 Zobrazit plný text záznamu