Performance Evaluation of Principal Component Analysis for Clustering on Sugarcane Dataset

Principal Component Analysis (PCA) is a tool which is used to reduce the high dimensional dataset to lower-dimensional dataset without losing the information from it. It is usually advisable to apply PCA on the multivariate dataset to get more accuracy in the formation of the clusters, before applying the clustering techniques of Data Mining. The experimentation is done for delineating the management zones in the field for the site-specific management of the nutrients. Soil samples of the Sugarcane field are taken as a datastet for the experimentation. The authors have tried to evaluate the impact of application of PCA technique versus no application of PCA technique on the formation of management zone in the undertaken field. The results indicate that, there is no significant impact of application of PCA in the clustering techniques, on the formation of management zones. The nutrient distribution across zones formed, is not affected by the application of PCA technique.