Modelo de inteligencia computacional para la determinación del grado de afectación de la roya en café robusta

This research was proposed with the purpose of determining the degree of affectation of rust in robusta coffee plants through the implementation of a computational intelligence (CI) model. To do this: (1) a bibliographic review was carried out to obtain information about the models that are used in research related to the object of study, concluding that the Support Vector Machines (SVM) is the most used and presents better results; (2) with the data set provided by Agrícola's career, made up of 12 phenotypic characteristics of coffee specimens affected by rust, several versions of data sets were constructed using 3 oversampling criteria; (3) the SVM model was implemented, which was trained and tested with each of the versions of the data sets created, (4) later, the use of IC model validation techniques They allow to make adjustments on the model until obtain results that reflect a high degree of precision by making inferences. This process that the definitive model obtained a performance of 80% during the tests and a 74% precision (on average) during the validation, therefore, a model was created that registers a relatively high performance that can also be used in other research related to combating coffee rust disease.

• Հիմնական հեղինակ: Andrade Sánchez, Cesar Israel (author)
• Ձևաչափ: bachelorThesis
• Լեզու: spa
• Հրապարակվել է: 2021
• Խորագրեր: Inteligencia computacional; Café robusto; Roya; Validación cruzada
• Առցանց հասանելիություն: http://repositorio.espam.edu.ec/handle/42000/1575