Determination of Factors Affecting Hazelnut Farmers' Agricultural Insurance by Data Mining Algorithms

Abstract

Agriculture is an open factory. With this characteristic of it, it faces many risks and uncertainties at any time. Farms are faced with natural, social and economical risks during the agricultural production period. Risk arising from natural conditions is the one of the most important risks in the agricultural sector. Agriculture is the only sector that meets the nutritional needs of the society. So it is very important protection of agriculture against risks. Agricultural insurance is one of the primary measures taken by farmers against these risks and uncertainties. For this reason, it is of great importance for farmers to have agricultural insurance in order to ensure economic sustainability. The main material of the study is the data collected from 70 different hazelnut producers in 7 villages in the Karasu district in the city of Sakarya in Turkey and they were selected with the purposeful sampling method. In the current study, it is aimed to determine the factors affecting farmers' getting agricultural insurance. According to the decision tree model created in the current study, it was determined that the factors affecting farmers' getting insurance are the amount of hazelnut production, non-agricultural income status, farmer's agricultural experience, total agricultural land assets and the profitability of hazelnut production. In the current study, the performances of different classification algorithms were also compared. According to the results of the research, it was determined that the classification algorithms used gave successful results. According to the results obtained with percentage split method, J48 and PART algorithms were determined to have the highest degree of accuracy in the cross validation method.