Development of Decision Support System Based on the Bayes ARD Algorithm for Irrigation of Cotton

Abstract

Cotton is a plant, which is mainly cultivated in regions where the irrigation is necessary as rainwater is not adequate. Researches in the recent years have showed that the irrigation water used could be declined. Improvements in the technological field has made Decision Support Systems combined with Neural Networks and data analysis, an important tool of sustainable agriculture. Cotton producers need to reduce irrigation water needs and that can be achieved by using new technologies. The development Decision Support System was conducted, having 3 different types of input. Data derived from a variety of IoT sensors, weather stations, and on-site measurements (yield and ΕΜ38) derived from 3 fields in Greece, creating a dataset of 9 different inputs. A total of 13 different algorithms were tested and evaluated in order to determine which one is the ideal for our dataset. The adoption of this technology in real data predicted the reduction of the irrigation times, ensuring that there will be no losses in the final yield.