The Novel Technique Based Gaussian Firefly Selection with Modified Fuzzy Cognitive Maps Prediction on Vegetable Plants.

Abstract

The broadest economic zone in developing countries is agriculture and it has the major responsibility for improving the economic growth of the country. Over the past decades, the growth of agriculture is monitoring by using the different data mining techniques. However, the plant yield prediction requires further improvement. Hence in this article, the plant yield prediction is improved by considering the different features such as soil and weather characteristics. Such features are extracted and the most optimal characteristics are chosen by using the Non-Dominated Sorting Firefly (NDSF) and Gaussian Firefly (GF) algorithm. These algorithms solve the Pareto-front issue and movement speed of firefly towards the best global solution. Then, the selected features are classified by using the Modified Fuzzy Cognitive Map (MFCM) algorithm for predicting the growth of plant yield. Finally, the predicted outcomes are broadcasted to the farmers for identifying the causes for plant yield degradation. The experimental results illustrate that the proposed GFFS-MFCM based plant yield prediction achieves high accuracy compared with the other techniques.