Welcome to the InfoSci Platform
Reference Hub15
Two-Level Classifier Ensembles for Coffee Rust Estimation in Colombian Crops

Two-Level Classifier Ensembles for Coffee Rust Estimation in Colombian Crops

David Camilo Corrales, Apolinar Figueroa Casas, Agapito Ledezma, Juan Carlos Corrales
Copyright: © 2016 |Volume: 7 |Issue: 3 |Pages: 19
ISSN: 1947-3192|EISSN: 1947-3206|EISBN13: 9781466692008|DOI: 10.4018/IJAEIS.2016070103
Cite Article Cite Article

MLA

Corrales, David Camilo, et al. "Two-Level Classifier Ensembles for Coffee Rust Estimation in Colombian Crops." IJAEIS vol.7, no.3 2016: pp.41-59. http://doi.org/10.4018/IJAEIS.2016070103

APA

Corrales, D. C., Casas, A. F., Ledezma, A., & Corrales, J. C. (2016). Two-Level Classifier Ensembles for Coffee Rust Estimation in Colombian Crops. International Journal of Agricultural and Environmental Information Systems (IJAEIS), 7(3), 41-59. http://doi.org/10.4018/IJAEIS.2016070103

Chicago

Corrales, David Camilo, et al. "Two-Level Classifier Ensembles for Coffee Rust Estimation in Colombian Crops," International Journal of Agricultural and Environmental Information Systems (IJAEIS) 7, no.3: 41-59. http://doi.org/10.4018/IJAEIS.2016070103

Export Reference

Mendeley
Favorite Full-Issue Download

Abstract

Rust is a disease that leads to considerable losses in the worldwide coffee industry. There are many contributing factors to the onset of coffee rust e.g. Crop management decisions and the prevailing weather. In Colombia the coffee production has been considerably reduced by 31% on average during the epidemic years compared with 2007. Recent research efforts focus on detection of disease incidence using simple classifiers. Authors in the computer field propose alternatives for improve the outcomes, making use of techniques that combine classifiers named ensemble methods. Therefore they proposed two-level classifier ensembles for coffee rust estimation in Colombian crops using Back Propagation Neural Networks, Regression Tree M5 and Support Vector Regression. Their ensemble approach outperformed the classical approaches as simple classifiers and ensemble methods in terms of Pearson's Correlation Coefficient, Mean Absolute Error and Root Mean Squared Error.

Request Access

You do not own this content. Please login to recommend this title to your institution's librarian or purchase it from the IGI Global bookstore.