Principal Component Analysis (PCA)

Back to Index

---

Objective

Learn PCA

Prerequisite Reading

• Brush up on Unsupervised learning

About Dimension Reduction

• Understanding dimensionality reduction
• A beginner’s guide to dimensionality reduction in Machine Learning
• A cool video explaining dim reduction - must watch this video for the special effects and you too will become a fan of Siraj Raval
• Dimensionality Reduction - a good video intro

PCA - Principal Component Analysis

• Principal Component Analysis for Dimensionality Reduction in Python - good intro and code
• Introduction To Principal Component Analysis In Machine Learning
• PCA using Python (scikit-learn)

Extra Reading

A Little Math

• A very nice explainer video of Eigenvectors from 3 Blue 1 Brown - I am fan !
• Eigenvalues and Eigenvectors

Knowledge Check

• Why not use all dimensions for ML?
• What is the difference between feature selection and dimension reduction?
• What is the use case for PCA?
• What are principal components (PC)?
• What would be the trend of successive eigen values of PCs?
• What about cummulative eigen values of PCs?
• What does a correlation matrix of PCs look like?

Exercises

Difficulty Level

★☆☆ - Easy
★★☆ - Medium
★★★ - Challenging
★★★★ - Bonus

EX-1: Using PCA to visualize (★☆☆)

Start with this pca-1-intro notebook.

Here we will reduce dimensions of a mtcars dataset to 2 dimensions so we can do a plot

EX-2: PCA on wine quality data (★★☆)

Start with this pca-2-wine-quality notebook.

We will perform PCA wine quality data

More Exercises

---

Index