A New Paradigm for Distributed Data Analysis in Terra Scale
Vithanage Pemajayantha · International journal of tomography and simulation · 2010
A major problem in distributed data mining is the network traffic, and a solution to that is found with a new paradigm in which distributed data analysis is carried out in abstract forms of partial analysis of raw data (‘paragrams’) which are transmitted through a network in place of large terra-scale data sets thus reducing network traffics. With the new paradigm, any statistical analysis is possible and this possibility is illustrated with examples of simulated data sets. Methods discussed here have straightforward applications in distributed data analysis in terra scale.