Dimensionality Reduction Procedure for Bigdata in Machine Learning Techniques

Abstract

In the present field of software applications, the prominently employed parameters for parameters control are the kinds of models such as cloud computing, machine learning, and big data analytics. So, in the current scenario, these are in high demand and are on-line with the trends for future decades as well. Nevertheless, as mentioned earlier, these models can access very low data and process speed. It is well known that the storage equipment’s for day-to-day monitoring serves at a higher cost and has hardware complexity, further leading towards rapid increment in dimensionality. Therefore, for the higher rate of dimensional data, the optimization approach of any variety would consume time to a greater extent. The concern issues are mostly related to the dimensionality with high data space instead of the low data space. A dimensional dropped approach is proposed in this paper in combinational with the Logistic regression (L.R.) version. The proposed technique is well known and applicable for the problems of clustering and dimension reduction. The size of the dimensional data to the LRML method has diminished, and the efficiency achieved at the rate of 95.5% and the reduction ratio is 34.89%.