International Journal For Multidisciplinary Research

E-ISSN: 2582-2160     Impact Factor: 9.24

A Widely Indexed Open Access Peer Reviewed Multidisciplinary Bi-monthly Scholarly International Journal

Call for Paper Volume 7, Issue 4 (July-August 2025) Submit your research before last 3 days of August to publish your research paper in the issue of July-August.

Pearson’s Correlation in Predictive Analytics and Machine Learning: Applications & Limitations

Author(s) Jarita Das
Country India
Abstract Pearson’s correlation coefficient serves as a vital statistical measure in predictive analytics and machine learning by offering profound insights into linear relationships between variables. It is, thus, instrumental in understanding variable relationships, selecting features, detecting multicollinearity, and assessing the model performance. This paper explores the applications of Pearson’s correlation in selecting features, reducing dimensionality, and interpreting the selected model. The paper highlights importance of Pearson’s correlation in identifying suitable predictors and improving algorithmic performance in predictive analytics and machine learning. The paper also takes into account the limitations of Pearson’s correlation that includes its sensitivity to outliers and reliance on assumptions of linearity and normality at the exclusion of non-linear associations. Alternative correlation measures like Spearman’s rank and mutual information that address the shortcomings of Pearson’s correlation are also taken within the purview of discussion. Through an examination of both the strengths and weaknesses of Pearson’s correlation, the paper sheds light into use of Pearson’s correlation in predictive modeling while stressing the need for adhering to complementary techniques in advanced machine learning applications.
Keywords Pearson’s correlation, predictive analytics, machine learning, Spearman’s rank, Kendall’s Tau
Field Mathematics > Statistics
Published In Volume 5, Issue 3, May-June 2023
Published On 2023-05-06
DOI https://doi.org/10.36948/ijfmr.2023.v05i03.48969
Short DOI https://doi.org/g9qxdj

Share this