Polyanalyst application for forest data mining
Carole MAI, M. Krishna, Annapareddy V. N. Reddy · 2005
Extraction of implicit knowledge and spatial relationship is one of the principal applications of data mining. Polyanalyst is a user friendly software package for spatial data mining. It has package modules to handle association rules, classification and prediction and cluster analysis. In this study data mining is applied to forest data to determine the factors responsible for deforestation. Four major factors - roads, villages, human population and cattle are taken into consideration to assess the extent of deforestation. Satellite images of forest area along with collateral data are the basic input for this project. The linear regression module is used for prediction. Identification of attributes which are to be included in the exploration and targeted for prediction is felicitated through Polyanalyst's multi-parametric stepwise linear regression modules. The predicted vs real graph shows the points in the actual data set, being explored along with where these data points would have been predicted to fall by the model produced. The study is an initial attempt to assess the scope of Polyanalyst for forest data mining.