Para computing for distributed data analysis in terra scale

![CDATA[It is well known that one of the major constraints in distributed data mining is the network traffic, and here we proposed a valuable solution to that problem using repeated parameter estimation method. Para Computing in Distributed data mining is a new concept in that partial analyses of the raw data are mined from the network so that network traffic is minimized within a framework of real time data entry and ‘commutative analysis’. Here we refer commutative analysis as a technique that partial data analysis is carried out within small segments of micro-network in real time with an expected mining at a later stage in a large terra-scale network. The procedure packed with commutative analysis is referred to as Para Computing, and it is possible with repeated parameter estimation methods. This possibility was investigated thus resulting in a new paradigm for terra scale data mining.]]