Classification of soybean pest data using decision tree algorithm

Classification of large volume of data especially in agriculture is a challenging task. Decision tree method isgenerally used for the classification, because it is the simple hierarchical structure for the user understanding and decision making. In the present study, the various classification techniques have been applied with Spodoptera spp. solitary larvae data set ofsoybean, for classifying into four classes based on Economic Threshold Level (ETL), using R statistical language. Out of six classification methods tested, it was found that C4.5 (decision tree) was effective with accuracy of 78 per cent followed by Naïve Bayes and kNN algorithms both with 72 per cent accuracy.