Ciao, I have to perform an estimation of price fluctations. There are some outliers in the price, namely above 350€ and above 950€. My first feeling is to add two booleans representing prices between 350 and 950 and above 950 to correct the intercept. Would that be ok? The model hoops from R-square 0.44 to 0.77 but I'm wondering if that's mathematically correct.
Asked
Active
Viewed 8 times
0
-
No, this isn't a good method. See this thread https://stats.stackexchange.com/questions/175/how-should-outliers-be-dealt-with-in-linear-regression-analysis for some good ways to deal with outliers in regression. – Peter Flom Jul 29 '23 at 11:52