Crop recommendation system and pest classification using weighted support vector machine on climate data

Abstract

Introduction: the primary cause of the significant decline in crop productivity is farmers' poor crop selection. A number of pests, including weeds, insects, plant diseases, and the poisonous nature of the most current remedies, offer challenges to the current approach. Therefore, for the most effective and precise classification and recommendations, these factors should be considered together. Methods: levy flight Grey Wolf Optimization (LGWO) and the WSVM (Weight-Support Vector Machine) method are recommended in this research for the intention of upgrading the efficiency of the system as well as resolving the above-mentioned issues. A CRS (Crop Recommendation System) utilizing the LGWO-WSVM algorithm is to be developed in order to increase crop productivity. This study's primary stages include crop suggestion, FS (Feature selection), and pre-processing. The KNN (K-Nearest Neighbour) technique is utilized for the pre-processing of the climatic dataset in order to accommodate incorrect values and missing variables. Results: the best fitness values are utilized to identify more pertinent weather features. These chosen qualities are then applied to the categorization phase. In order to create a system which integrates the predictions of the LGWO-WSVM model to recommend an appropriate crop depends on the kinds of the particular soil and features having greater accuracy. Conclusion: in order to get the best recommendation outcomes, it is also utilized to categorize the pest traits. The test outcomes indicate that the recommended LGWO-WSVM strategy overtakes the current methods by accuracy, precision, recall, and execution time