DefinePK

Abstract

In India, agriculture is considered to be a prime activity for the most of populations. Therefore, an economic growth a nation mainly depends on the development of agricultural activities like improving crop production, utilizing developed technologies to monitor crop yield, etc. As a result, different crop yield monitoring systems have been developed to enhance the agricultural productivity. Among different systems, multi-objective firefly Optimized Fuzzy Cognitive Map (OFCM) was proposed for predicting the Arachis Hypogaea (groundnut) yield by using both soil and weather factors. Here, multi-objective firefly was applied for learning FCM by optimizing the weight parameters utilized in FCM. However, the Pareto-front issue has occurred in the firefly algorithm due to consider the multiple objective functions. Hence in this article, crowding distance between fireflies is computed for choosing appropriate fireflies. Moreover, an Improved Optimized OFCM (IOFCM) is proposed in which a modified multi-objective firefly optimization is used to minimize the randomness and achieve the global optima by improving the movement of fireflies. Though it achieves better optimization, FCM has high sensitive while input data are missing resulting in prediction decision is made with incomplete information. As a result, a modification in FCM is proposed to improve the prediction performance more effectively. In this modification, the value of each node in the FCM is computed by considering the hesitancy function that increases the prediction accuracy even some input data are missed. This newly proposed algorithm is called an Improved Optimized Intuitionistic FCM (IOIFCM). Finally, the experimental results show that the effectiveness of the proposed IOIFCM based crop yield prediction compared to the other optimization algorithms.