Prediction of The Factors Affecting Watermelon (Citrullus lanatus (Thunb.) Matsum. & Nakai) Yield Using Data Mining

Abstract

The aim of this study was to evaluate the factors of affecting watermelon yield in Diyarbakır province. The data was obtained from surveying of 80 watermelon farmers in Diyarbakır province, Turkey by Simple Random Sampling Method using the Chi-square automatic interaction detector (EXHAUSTIVE CHAID) algorithm of the Data Mining Regression Tree methods. In the model created, the dependent variable was WY (watermelon yield), and the independent variables were determined as R (region), AF (age of farmer), EL (education level), CA (cultivation are), FD (fertilization date), FA (amount of fertilization), DS (date of spraying), AS (amount of spraying), NI (number of irrigation), IT (irrigation time), AN (anchor number), HT (harvest time). As a result of the study, the factors that significantly affect the yield of watermelon; AN, NI, HT, CA, R has been determined. An average of 4488.9 kg watermelon yield per decare was obtained and the number of hoes was the variable that most affected the watermelon yield. As a result in order to get a higher yield per unit area, watermelon producers should anchor number more than 4 times, irrigate 5 to 6 times at less than 2 hours, and apply fertilizer in May. In addition, Çermik, Eğil, Yenişehir and Bismil were determined as more suitable regions for watermelon production.