Principal Component Analysis (PCA) and Factor Analysis

Analytics / Machine Learning / Dimensionality Reduction : PCA & Factor Analysis using SAS and R program

The course explains one of the important aspect of machine learning – Principal component analysis and factor analysis in a very easy to understand manner. It explains theory as well as demonstrates how to use SAS and R for the purpose.

What you’ll learn

  • Understand Principal Component Analysis and Factor Anallysis in crysal clear manner.
  • Will know how to coduct principal component analysis and factor analysis using SAS / R.
  • Will understand, how PCA helps in dimensionality reduction.
  • Will understand the difference and similarity between PCA and factor analysis.
  • Students will be able to use PCA for variable selection.

Course Content

  • Principal Component Analysis (PCA) –> 9 lectures • 54min.
  • Factor Analysis –> 4 lectures • 34min.
  • Using Principal Component Analysis for Variable selection –> 5 lectures • 12min.

Principal Component Analysis (PCA) and Factor Analysis

Requirements

  • The course will start with elementary concepts but knowledge of basic statistics will help.
  • For execution – it will help to know basic SAS or R programming.

The course explains one of the important aspect of machine learning – Principal component analysis and factor analysis in a very easy to understand manner. It explains theory as well as demonstrates how to use SAS and R for the purpose.

The course provides entire course content available to download in PDF format, data set and code files. The detail course content is as follows.

  • Intuitive Understanding of PCA 2D Case
    1. what is the variance in the data in different dimensions?
    2. what is principal component?
  • Formal definition of PCs
    1. Understand the formal definition of PCA
  • Properties of Principal Components
    1. Understanding principal component analysis (PCA) definition using a 3D image
  • Properties of Principal Components
    1. Summarize PCA concepts
    2. Understand why first eigen value is bigger than second, second is bigger than third and so on
  • Data Treatment for conducting PCA
    1. How to treat ordinal variables?
    2. How to treat numeric variables?
  • Conduct PCA using SAS: Understand
    1. Correlation Matrix
    2. Eigen value table
    3. Scree plot
    4. How many pricipal components one should keep?
    5. How is principal components getting derived?
  • Conduct PCA using R
  • Introduction to Factor Analysis
    1. Introduction to factor analysis
    2. Factor analysis vs PCA side by side
  • Factor Analysis Using R
  • Factor Analysis Using SAS
  • Theory for using PCA for Variable Selection
  • Demo of using PCA for Variable Selection