Answer ( 1 )

  1. In linear regression, one of the basic assumption is there should not be influential variables or outliers as the presence of outliers can decrease the accuracy because regression is sensitive to outliers. So to deal with these variables we can replace them with the most appropriate measure (it can be mean, median or mode) depending on the type of distribution they have.
    Definately it will be found useful.

    Best answer

Leave an answer

Browse
Browse