Clustering Algorithms and Bayesian Networks for Distributed Geospatial Data Mining and Knowledge Discovery
Italian National Research Council, Italy.
A few consolidated methods of data mining approaches developed by ourselves are proposed in the modern framework of geoinformatic remote sensing. These approaches and combinations of them, be it partially or fully, helps in extracting knowledge from huge data sets especially as in geo-informatics. These methods of data mining are quite useful in depicting trends and patterns associated with huge amount of partially correlated data generated at various stations and classifying them based on different variables associated with the process which can also be in a non-linear fashion. These methods are successfully applied individually in various contexts. We suggest that combinations of these approaches when worked upon yield an effective classification of data even in the complicated and distributed field of geo-informatics.
Key words: Bayesian Networks, Unsupervised clustering, Data mining, Algorithms.