11-24-2016 11:39 AM
Can anyone help me to identify how can I predict the most influential variable impacting the model in linear regression.
Which statistic or method I need to refer for this
11-25-2016 01:33 AM
Thanks for the reply. You mean smallest value or largest t value would have greatest impact or significance to the dependant variable. What about coeff var?
11-25-2016 05:50 AM
No. The coefficient of variable only measure how much Y will change when X change one unit. Y(x+1)-Y(x)=beta(x+1)-beta*x=beta It doesn't measure the importance of variable to the model.
Need further help from the community? Please ask a new question.