Statistical Methods in the Usage of Correlation and Regression of the Machine Learning Models
V. Manikyala Rao1, K. Uma Pavan Kumar2, B. Priyanka3
1V. Manikyala Rao, Assistant Professor, Department of CSE, Malla Reddy Institute of Technology, (Telangana), India.
2Dr. K. Uma Pavan Kumar, Associate Professor, Department of CSE, Malla Reddy Institute of Technology, (Telangana), India.
3B. Priyanka, Assistant Professor, Department of CSE, Malla Reddy Institute of Technology, (Telangana), India.
Manuscript received on 15 October 2019 | Revised Manuscript received on 24 October 2019 | Manuscript Published on 02 November 2019 | PP: 2489-2491 | Volume-8 Issue-2S11 September 2019 | Retrieval Number: B12930982S1119/2019©BEIESP | DOI: 10.35940/ijrte.B1293.0982S1119
Open Access | Editorial and Publishing Policies | Cite | Mendeley | Indexing and Abstracting
© The Authors. Blue Eyes Intelligence Engineering and Sciences Publication (BEIESP). This is an open access article under the CC-BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/)
Abstract: The predictive analytics is the most commonly used methodology in the usage of Machine Learning class of algorithms. Based on the values generated at the time of running the algorithm the significance of the model can be estimated. The current work gives a complete focus on P value and the significance levels of the P value in the correlation analysis of the algorithms. Based on the P value the impact of the model can be notified and the interpretation of the results can be done in the efficient way. The other dimension of the work is the usage of statistical functionalities in the regression analysis, most of the researchers are focusing on the shallow usage of regression analysis in the classification of the tasks. The current work explains the complete internals of the regression models available and the usage of the statistical functionalities utilized in the implementation of the corresponding variants of the algorithms. We believe that the current work exclusively helps the upcoming researcher in the areas of regression in the context of the statistical functionalities which are vital in the implementation of the tasks. The outcome of the work is to exploit the correlation analysis with various significance levels and the issues in the processing of the analysis. The another point here is the regression internals with the focus of statistical methods available in the processing of regression variants. The regression analysis involves various types like linear regression, multiple regressions and logistic regression. The current work gives an overview of all these three types of regressions and also the significance of P value in the prediction of outcome. In the examples such as house rate prediction based on the given area, salary of an employee based on the experience level, profit of the start-up companies based on the spending on research, admin marketing and state of the country are best suitable in the explanation of regression.
Keywords: Regression, Statistical Methods, P-Value, Correlation.
Scope of the Article: Machine Learning