Coffee Farming, Machine Learning, Water Potential, Data Analysis, Reflectance
Abstract
Water potential is an important indicator used to study water relations in plants, as it reflects the level of hydration in their tissues. There are different numerical variables that describe plant properties and can be acquired from leaf reflectance. In this study, the objective is to explore spectral variables to estimate water potential in coffee trees, using computational intelligence tools. Furthermore, the data present two crop management groups, irrigated (subjected to irrigation methods) and rainfed (coffee plants were not exposed to artificial irrigation). Four Machine Learning techniques were implemented: MLP-type Artificial Neural Network (Multi-Layer Perceptron), Decision Tree, Random Forest and KNN (K- Nearest Neighbor). Two distinct methods were implemented for the four techniques, estimation and classification. The results show that artificial neural networks were superior for both estimation and classification approaches.
