Incremental algorithm for Distributed Data Mining

The necessity of computational grid is driven by large volumes of data. Such grids are designed giving more emphasis on the efficiency of computation. Data set is divided into a number of sets and sent to a number of processing data nodes. This way of dividing the data sets for parallel computing is heavily applied in grids. But only dividing the datasets is no way, in the real sense, an improvement in efficiency. Unless we have efficient algorithms to process the data, just applying the strategy of computational parallelism will only ask for more and more computational resources. Here we present one such algorithm for the task of data classification.