Welcome to the homepage of Math F614 Numerical Linear Algebra in the Dept. of Mathematics and Statistics at the University of Alaska Fairbanks.

Sections

If you plan to be present on campus in Fairbanks during the semester, please sign up for the in-person “901” section (crn 75364), and plan to attend lecture in Chapman 107. If you are remote, signing up for the web-based “701” section (crn 75371) is just fine, and you should see the Canvas page for the Zoom link.

Instructor: Ed Bueler

Email me at elbueler@alaska.edu. I hold office hours in Chapman 306C.

Canvas course page

Log in to canvas.alaska.edu/courses/27130 for the lecture Zoom link, Homework and Exam solutions, and to see your grades.

Getting Started

• Attend lectures: MWF 2:15-3:15pm in Chapman 107, or online (see Canvas page for Zoom link).

• Read the Syllabus (PDF).

• See the Schedule (PDF). Check it often for due dates, and for which topics come next!

• Check out the nearly-weekly homework Assignments.

• There are three Exams, namely two Midterm Quizzes and a Final. All are in class, and the Final will happen at the scheduled time. See the Exams tab for review guides and sample exams.

• Homework will require programming in Matlab/Octave or some other suitable scientific computing language. See the this page and the Programming languages compared (PDF) handout for recommended languages. Code example are at the Codes tab.

• A very good free online textbook makes sense as a supplement: Advanced Linear Algebra: Foundations to Frontiers. Each section has a short online video.

• What are we studying? Check out these Wikipedia pages for the topics we will get to, in this approximate order:

  • matrix-vector and matrix-matrix multiplication
  • orthogonal matrices
  • singular value decomposition
  • QR factorization
  • least squares methods
  • over- and under-determined linear systems
  • conditioning of problems
  • backward stable algorithms
  • Gaussian elimination (LU factorization) and its variants
  • Cholesky factorization
  • power iteration for eigenvalues, an important and impractical method
  • QR algorithm for eigenvalues, very practical
  • Krylov subspace iterations

---

Site design derived from coordinated Calc I, an original Jekyll design by David Maxwell.