Geometric interpretation of a correlation

Abstract

The study shows that the Pearson’s coefficient of correlation is equivalent to the cosine of the angle between random variables. It was found that the information about the intensity of the relationship between variables is included in the value of the angle between random vectors. The paper proposes intuitive criteria for measuring the intensity of the relationship between random variables.

Keywords

• data mining
• correlation coefficient
• cosine
• angle between vectors
• stochastic dependence