BITL2 - Introduction to Linear Algebra - Spring 2022

Other offerings of this course: Fall 2021, Spring 2022, Spring 2023

Time: Mondays and Wednesdays, 14:00 - 15:40
Instructor: Jānis Lazovskis, janis.lazovskis@rbs.lv
Office: 404, Skolas iela 11
Office hours: Every day, bookable online at calendly.com/jlazovskis
Internal page: ORTUS e-studijas

Course material

Syllabus and calendar of the course
Lecture notes (updated weekly)
Textbook (Strang's "Introduction to Linear Algebra")
Whiteboards for in-class collaborative work:
- Board 1 (Jan 3 - Feb 7)
- Board 2 (Feb 7 - Mar 9)
- Board 3 (Mar 9 - Apr 6)
Poster (A2 size) of all the topics discussed this semester, and their relations with each other

Homework assignments

Assignment 1: Due at the start of class Monday, January 17
Assignment 2: Due at the start of class Wednesday, February 2
Assignment 3: Due at the start of class Wednesday, February 23
Assignment 4: Due at the start of class Monday, March 14
Assignment 5: Due Wednesday, April 13, by midnight

Python and Mathematica code

Lecture 8: Orthogonality
- Python notebook: for inspecting the orthogonality of random vectors
Lecture 11: The least squares approximation
- Python notebook: for plotting points, computing least squares approximations, and plotting the appropriate polynomials
Lecture 13: Generalized distances
- Python notebook: for generating random points, visualizing them, creating the associated distance matrix, and its dendrogram
Lecture 16: Eigenvalues and eigenvectors
- Jupyter notebook: for visualizing what eigenvalues and eigenvectors do to vectors in the plane
Lecture 21: Singular value decomposition
- Python notebook: for visualizing the singular value decomposition of grayscale images
Lecture 22: Principal component analysis
- pca-augstskolas: data set in CSV or Python pandas PKL of university instuctors and students in 2020 (source: Oficiālais Statistikas Portāls)
- Least squares / PCA comparison: Two interactives to compare the difference between least squares (minimizing vertical distance) and the first principal component from PCA (minimizing perpendicular distance)