Improved Soil Data Prediction Model Base Bioinspired K-Nearest Neighbor Techniques for Spatial Data Analysis in Coimbatore Region

Abstract

In this research paper, agricultural Data Mining data are summarized. An improved Soil Data Prediction Model is developed to estimate the above parameters at locations for Coimbatore city. 142 locations were investigated for the development of the model. The model involves multiple regression equation, chi-square test and Bio inspired k-nearest neighbor classification. The GIS was used to manage the database and to develop thematic maps for depth, N value, free swell, liquid limit, plastic limit, plastic index, percentage gravel, percentage sand and percentage slit and clay. Field and laboratory studies were conducted in four locations and are compared with the predicted values.