STAT 154 Lecture Notes - Lecture 18: Simple Linear Regression, Botulinum Toxin, Data Set
25 Oct 2018
School
Department
Course
Professor
Document Summary
High how many predictors dimension data y to fix ) find. A ? p - predicts m predicts my principle partial h > m component regression least square regression. Z ) when we use the ( xt x ) principle component n a. Z regression , what"s missing ? cuseful the what. How of to calculate transformation responsive matrix information in. Arg xp information on the input and. Partial least squares approach: identifying m new features using the following partial method: after standardizing the p predictors, nd 1 by setting each coe cient 1 to be the coe cient of simple linear regression of y onto. ; nd 2 by regressing each variable on 1 and then take the residuals, nd. 2 using the orthogonalized data in the same way as 1was constructed using the original data; using the same fashion to nd all. A data set for . fit c salary -
