Feature Selection in Multiple Linear Regression

AD2019AD2019 Member, University ProfessorPosts:13University Professor
edited October 2019 inHelp
I selected "t-test" as the feature selection method in a multiple regression model, and set alpha to 0.01. Why is the regression output including an independent variable with a p-value of 0.05? Only when I reduce the alpha to 0.001 does this variable go away. If I set alpha = 0.01, shouldn't the selected model only show me independent variables whose p-value is less than 0.01?
thank you in advance for answers.
AD
Tagged:

Answers

  • sgenzersgenzer Administrator, Moderator, Employee, RapidMiner Certified Analyst, Community Manager, Member, University Professor, PM ModeratorPosts:2,959Community Manager
    嗯@AD2019let's see if someone like@mschmitz @David_A @IngoRMcan help here.

    Scott
  • Telcontar120Telcontar120 Moderator, RapidMiner Certified Analyst, RapidMiner Certified Expert, MemberPosts:1,635Unicorn
    Can you be more specific about what you are experiencing or provide sample xml or output?
    Do you have any attributes that are generated from recoding numerical to nominal? I think RapidMiner might keep those as a group if so which means some of the individual values have p values greater than the threshold but the overall attribute does not, although it would be good for one of the RM staffers to confirm that.
    Brian T.
    Lindon Ventures
    Data Science Consulting from Certified RapidMiner Experts
Sign InorRegisterto comment.