AnalyticBridge
A Data Science Central Community
Prediction using linear regression
Hi Everyone,
Prediction of count(bike renting):
I have 12 variables, out of which one is dependent variable which should be predicted from others. I implemented Step-wise for selection of variables in the equation using R-tool. But what i get maximum adjusted R-square is 45% -- outliers are removed. Could any suggest me how to improve my model?
Thanks!
Tags: modelling, predictive
Replies to This Discussion
-
Permalink Reply by KOUADIO Michel on
-
The explanatory power of your model (45%) is very weak.
It means that the relation between variables is not probably linear.
Thus you have to use another technique of modelling
-
Permalink Reply by Ved on -
Sometimes Stepwise regression is bad idea for modelling. In Stepwise regression, adjusted R2 values might be high in start, and then dip sharply as the model progresses. If this happens, identify the variables that were added or removed when this happens and adjust the model.
If this problem still persist, then look for other modelling techniques.
-
Permalink Reply by Sagar Diwakar Uparkar on
-
Hi Rupesh,
Good evening. The prediction ability of your model is still good as in real life this is the only chances of getting adjusted R squre.
Apart from this what you can check, is there any multicolinearity exist which may drop your adjusted R squre.
© 2019 AnalyticBridge.com is a subsidiary and dedicated channel of Data Science Central LLC
Powered by
Badges | Report an Issue | Privacy Policy | Terms of Service
