An efficient approach for iterative learning algorithms

Abstract

In this paper, a framework which takes into account machine learning for the analysis of massive datasets is proposed. The framework maps the algorithms to the respective platform so as to extract maximum resource efficiency. In addition, the framework takes into account a data projection technique called as Elastic Dictionary to form sparse representation of the underlying data. By this way, the resource efficiency is optimized leading to reduction in the cost associated with the performance. The framework represents a model and shows the performance metrics in accordance with their respective runtime and storage. An additional application program interface takes into account the applicability of the framework to the underlying platform or datasets. The framework is based on the union of both the content and platform aware methodologies so as to make the machine learning algorithms to utilize the resources efficiently.