Quantum-Behaved Bat Algorithm (QBBA) based feature selection for trade off between Weather-Soil-Yield relationships in rice using Data mining methods
Abstract: Fluctuations in the global production of major crops are important drivers of food prices, food security and land use decisions. Average total yields for these commodities are determined by the performance of crops in millions of fields distributed across a range of management, soil and climate regimes. These has been becomes a major important issues for farmers. Among them, climatic conditions and soil characteristics directly affect the performance of crops. To solve this problem the recent work introduces a new enhanced density based clustering algorithm for measuring weather soil- yield relationships. Feature selection is another part of work in weather soil- yield relationships. All of them are very important and challengeable. In this paper, mean max normalization, Z-Score Normalization methods and Quantum-Behaved Bat Algorithm (QBBA) are introduced first, then, discussed the relationship between feature selection and data normalization. This QBBA based feature selection algorithm selects most important features from the weather soil- yield characteristics. Then Parallel Enhanced Gaussian C Means (PEGCM) Clustering algorithm is introduced for the weather patterns corresponding to individual cropping events revealed different groups of weather patterns for irrigated and rainfed systems with contrasting yield levels. The results demonstrate that changes in the amount of soil and air temperature are the most representative weather predictors among the studied parameters. A new clustering algorithm was developed for the prediction of weather-soil- yield relationships and it is found that, under a specific amount of attributes, the risk of weather events in pollachi of Coimbatore region. The results show that the proposed PEGCM algorithm achieves better performance compared with the existing Enhanced DBSCAN (EDBSCAN), Density-based Spatial Clustering of Applications with Noise (DBSCAN), and Hierarchical Clustering (HC) in terms clustering accuracy, precision and recall.