Performance of Gradient Boosting Learning Algorithm for Crop Stress Identification in the Greenhouse Cultivation

Greenhouse cultivation is one of the most crucial circular economy systems in agriculture that allows high resource use efficiency and low environmental impact. The data generated in high-tech and sophisticated greenhouse operations are provided by a variety of different sensors that enable a better understanding of the operational environment. In this study a learning algorithm namely Gradient Boosting Machine was tested using a database that was generated in greenhouse experiments in order to estimate different types of stress in a tomato crop. The examined model performs qualitative classification of the data, depending on the type of stress (water, high or low temperature). For the comparison, a 10-fold cross validation strategy on the 10,763 samples from the training set was selected. The dataset was divided in two parts, one for training 80% (8,610) and a second one for validation 20% (2,152). The cross-validation process was repeated 50 times. Among the parameters used as input for model building, leaf temperature had the highest significance with a ratio of 0.51. According to the results, the Gradient Boosting algorithm defined all the cases with high accuracy. The model built could identify the complete set of 379 samples of plants with low temperature stress, the 1305 out of 1321 samples of the plants without stress and the 431 out of the 452 samples of the plants with water stress. The accuracy of the model in identifying the type of stress was 98%. This research is co-financed by Greece and the European Union (European Social Fund- ESF) through the Operational Programme «Human Resources Development, Education and Lifelong Learning» in the context of the project “Reinforcement of Postdoctoral Researchers - 2nd Cycle” (MIS-5033021), implemented by the State Scholarships Foundation (ΙΚΥ).