0
$\begingroup$

I am working on predicting BMI against weight, using linear regression.

The scatter plot of the data can be found below.

enter image description here

As you can see in the plot, there seems to be low (or no) correlation between the two variables and thus I have doubts whether I'm using the right method. Is it necessary to have correlated data in order to use linear regression? Would you advice me to try other methods or adding features?

Improve this question
$\endgroup$
6
  • $\begingroup$ Welcome to Data Science! What would “perfect correlation” look like to you? $\endgroup$
    – Dave
    Commented Sep 26, 2022 at 4:32
  • $\begingroup$ Isn't the function BMI(weight) already known? Why do you need to create a machine learning model instead of just using the formula? $\endgroup$
    – liakoyras
    Commented Oct 9, 2022 at 17:20
  • 1
    $\begingroup$ @liakoyras it looks like you didn't understand the question kindly $\endgroup$
    – tip. rock
    Commented Oct 10, 2022 at 7:51
  • 1
    $\begingroup$ @Dave, thanks so much, but I was looking for an answer kindly. $\endgroup$
    – tip. rock
    Commented Oct 10, 2022 at 7:52
  • $\begingroup$ Then what do you mean by perfect correlation? $\endgroup$
    – Dave
    Commented Oct 10, 2022 at 13:10

1 Answer 1

Reset to default
0
$\begingroup$

If there is no correlation between variables you cannot run regression analysis as one variable cannot predict the other. Hence, it is necessary to have correlated data in order to run linear regression.

Improve this answer
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.