A Practical Comparison on GIS Data of Two Data Mining Algorithms

This work explores the effectiveness of data mining classification techniques, their advantages and disadvantages. Classification is the largest of the applications, consisting in building models to predict belonging to a set of classes. In this study we compared using Weka tool two of the most known data mining algorithms on a collection of Geographic Information System (GIS), data called Cadastre which consist of a parcel plan from the Dolj area of Romania. From the performed experiments, results that the K-nearest neighbor algorithm works better that the Naïve Bayes algorithm in terms of accuracy.