Week 8

SVD and Applications

Important

• Session I Buffon (103A) Wednesday 13h30-15h
• Session II Sophie Germain (2012) Friday 9h00-10h30
• Calendar

Blackboard

• SVD
  • Definition
  • Existence
• Matrix norms
• SVD and low rank approximations
  • Eckart-Young-Mirsky theorem for operator norm
  • Eckart-Young-Mirsky theorem for Frobenius-Hilbert-Schmidt norm

Labs

• Lab 8 Principal Component Analysis as an application of SVD

We still rely on the previous labs

• Lab 1 - Introduction to R and RStudio
• Lab 2 - Introduction to Data Visualization using Gapminder dataset
• Lab 3 - Working with dplyr
• Lab 4 Univariate categorical variables
• Lab 5 Univariate numeric variables
• Lab 6 - Linear Regression I
• Lab 6 - Linear Regression II
• Lab 7 - Linear Regression II

Further work

Review the content of the two labs. Work out every part you do not already know. Report an issue if you are unhappy with the proposed solutions/hints.

Further reading

• Bin Yu and Rebecca Barter, Veridical Data Science
• R for data science
• Advanced R
  • S3 classes
  • S4 classes

Logistics

: you will work with the R programming language in this course.

You need either to install R, RStudio and VS Code on your computer, or to use your posit-cloud account.

Install Quarto on your computer to render the .qmd files.

Please follow the instructions here to install R, RStudio, VS Code, and Quarto or to access posit-cloud.

Please activate your ENT account (follow the instructions on Moodle). You will be able to access the PostGres server.

Back to Agenda ⏎