A Methodical Approach of Unlabelled Complex Big Data Analytics and Processing using Unsupervised Machine Learning Modeling

Abstract

Currently, unsupervised learning modeling and deep learning both envisioned for a better scope of knowledge extraction during big data learning scenarios. And mostly big data streams get generated from multiple sources and comprises of the hidden and unknown patter of information attributes, which requires efficient data learning mechanisms to be incorporated for a better scope of knowledge discovery. As big-data mostly contains unlabelled information, thereby extensive research effort has been laid towards applying unsupervised learning modeling. However, still, a gap exists in the conventional research approach in terms of complexity and learning time, which restricts their further case-studies into a sot-effective big-data analytics environment. To address this limitation, a theoretical design approach of heterogeneous big data learning is introduced in this study, which adopts a novel kernel oriented controller modeling (KOCM) approach to optimize the convergence performance of data learning within minimal computational steps. The performance of the approach is validated in terms of both computational efficiency and speed of computation. The outcome of the study shows KOCM outperforms the existing unsupervised approaches with a better scope of applicability into futuristic big data analytics systems.