Correlation is a measure of linear association between two variables X and Y, while linear regression is a technique to make predictions, using the following model:
Y = a0 + a1 X1 + ... + ak Xk + Error
Here Y is the response (what we want to predict, for instance revenue) while the Xi's are the predictors (say gender, with 0 = male, 1 = female, education level, age, etc.) The predictors are sometimes called independent variables, or features in machine learning.
Typically, the predictors are somewhat correlated to the response. In regression, we want to maximize the absolute value of the correlation between the observed response and the linear combination of the predictors. We choose the parameters a0, ..., ak that accomplish this goal. The square of the correlation coefficient in question is called the R-squared coefficient. The coefficients a0, ..., ak are called the model parameters and a0 (sometimes set to zero) is called the intercept.
There are various types of correlation coefficient as well as regression. For more details, see
Comment
Thank you Omer, I fixed the mistake.
Hi,
In the first paragraph, there is a mistake. In regression analysis, the dependent variable is denoted "Y" and the independent variables are denoted by "X". But it is stated here that " The predictors are sometimes called dependent variables, or features in machine learning.".
is it wrong to extend the scope of correlation to non-linear associations like exponential association between two variables for example x and y which are exponentially associated
© 2020 TechTarget, Inc. Powered by
Badges | Report an Issue | Privacy Policy | Terms of Service
Most Popular Content on DSC
To not miss this type of content in the future, subscribe to our newsletter.
Other popular resources
Archives: 2008-2014 | 2015-2016 | 2017-2019 | Book 1 | Book 2 | More
Most popular articles
You need to be a member of Data Science Central to add comments!
Join Data Science Central