Lecture 14 – Feature Engineering
Presented by Raguvir Kunani
Content by Raguvir Kunani, Joseph Gonzalez, John DeNero, Josh Hug
Note: In this lecture you see the
sklearn package being used to fit models. This video gives a quick guide on how the
sklearn package works.
A reminder – the right column of the table below contains Quick Checks. These are not required but suggested to help you check your understanding.
Motivating Feature Engineering
Applying nonlinear transformations to quantiative features. Incorporating domain knowledge.
Imputing missing values
Bag of words encoding. N-gram encoding.
Implications of feature engineering on normal equations