Framework for multi threads execution of data mining algorithms

Abstract

The present paper describes the framework for creating data mining algorithms from thread-safe functional blocks. This framework requirements decomposition of algorithms into independently functioning blocks. These blocks must have unified interfaces and implement pure functions. The framework allows create new data mining algorithms from existing blocks and improves the existing algorithms by optimizing single blocks or the whole structure of the algorithms. This becomes possible due to a number of important properties such as thread-safety inherent in pure functions and hence functional blocks.