Enhancing Open Cognitive Data Science Workbench on Open Power

Abstract

The current need for the Big Data that deals with structured and unstructured data, converts it into useful information, uses machine learning algorithms. The algorithms use statistical and mathematical problem-solving techniques. Much contribution has been made to mine data using machine learning algorithms, and analytical models are being built using the datasets. Hence, there is a need to create a standard platform to help the scientific community like academics, research scholars and organizations to install the scientific computing tools and to develop the cognitive application in a time-valued fashion. The objective of the proposed system is to improve the performance of machine learning algorithms and the data models; the data science workbench is being enabled on OpenPOWER. Building such a workbench on PowerPC will provide data and high-performance computing because of the significant advantage of Power on data workloads such as databases, data warehouses, data transaction processing, and indeed in high-performance computing. Parallelization of processes achieves this in power servers.