#linear-regression

Question 1: Select all that apply

mediumAI-checkedDimensionality Reduction

You are deciding whether PCA suits a dataset with many correlated numeric features, and you wonder how it relates to fitting a line. Which statements about PCA are true? Select all that apply.

Reducing $d$ dimensions to $k \le d$ with PCA means modelling the data with a $k$ -dimensional hyperplane1
PCA assumes that the variables in the dataset are uncorrelated2
You can tell which original features dominate a principal component by looking at its weights3
Since PCA and linear regression both minimize squared error, PCA is an unsupervised alternative to linear regression4

Timed test