OPTIMIZED DECISION TREE ALGORITHM WITH FRA FOR CROP YIELD PREDICTION IN IOT BASED AGRICULTURE

Abstract

In Internet of Things (IoT) based agricultural environment, crop yield prediction is one of the most desirable but difficult tasks for any country. Farmers are having troubles to achieve a good yield from crops because of unpredictable changes in the climate. To feed India's growing population, agriculture must incorporate cutting-edge technology and tools. Thus, machine learning techniques have been used widely to predict the crop yield. However, this study focuses to enhance the prediction accuracy using enhanced machine learning technique. Namely, for efficient crop yield prediction, an optimized decision tree (DT) classifier with fuzzy ranking algorithm (FRA) is used.  At first, FRA is presented to choose the features set from the crop yield dataset. Using the selected features, crop yield is predicted by presenting optimized DT. Using a modified penguin search algorithm (MPSA), the decision node of the DT is optimally chosen. The proposed prediction model improves the prediction accuracy.