From the course: R Essential Training Part 2: Modeling Data
Unlock the full course today
Join today to access over 22,600 courses taught by industry experts or purchase this course individually.
Predicting outcomes with quantile regression
From the course: R Essential Training Part 2: Modeling Data
Predicting outcomes with quantile regression
- [Instructor] One of the worst things you can have in data analysis is an outlier and when you're doing a uni-varied analysis looking at things like the mean, you know that a single outlier can arbitrarily distort or destroy the mean so that it doesn't tell you anything close to what you want. When it comes to associations like regression, outliers can be even worse depending on the nature of what's going on. And methods like least squares linear regression, that's the standard regression, can get thrown off hugely. Fortunately, there are ways to deal with this with robust methods. Robust means not as easily influenced by outliers and one of the best ways is with quantile regression. Think of that as analogy to the median as opposed to the mean but applied to the association of variables in regression. To show you how this works, I'm going to load a few packages including one called QuantReg for quantile regression.…
Practice while you learn with exercise files
Download the files the instructor uses to teach the course. Follow along and learn by watching, listening and practicing.
Contents
-
-
-
-
-
-
Predicting outcomes with linear regression8m 49s
-
(Locked)
Predicting outcomes with lasso regression7m 48s
-
(Locked)
Predicting outcomes with quantile regression6m 27s
-
(Locked)
Predicting outcomes with logistic regression12m 49s
-
(Locked)
Predicting outcomes with Poisson or log-linear regression3m 43s
-
(Locked)
Assessing predictions with blocked-entry models10m 35s
-
-
-