Numerical methods of linear algebra/A lineáris algebra numerikus módszerei

Fall 2017

In this semester the classes are in English.

Lecture: Mo,Tu 12:15-13:45,   IE 217.1

We are mostly following a Hungarian book but, for example, the book Matrix Computations by Golub and Van Loan is a good place to look for algorithms. Wikipedia also contains many useful things.

Main  topics covered. (There will be no class on the last week.)


always starting at  9:00. Let me know if none of them is good for you.

Steps before the exam:
  1. Choose one problem from this list (one that is not yet taken by someone else).
  2. Send me an email with your selection.
  3. If you are the first, it is yours (I ackwnowledge it).
  4. Solve it by a program you write or find somewhere (for example in MATLAB). If you really want, most of them can be solved with paper and pen only :)
  5. Play with it a little bit: use different precisions, add random noise or do some transformation  on the problem and see how this effects the solution; create larger example (randomly or by tensor product) and see how this effects the solution and the time it takes to compute, etc.
The exam consits of two parts.

In the first part you will talk about your experiences with the selected problem (method, result, other considerations). For this you can use whatever you wish: your notes, printout, computer.

The second part is an oral exam  from the material (definitions, algorithms, theorems, proofs, etc.)  covered during the semester. (There will be some time to prepare before you start talking about the topic selected for you.).
For this part, prepare  a "cheat sheet": a sheet of size A4, hand written by yourselves at home.

The topics for the second part are be listed here (there will be some time for preparation before you start talking about the one selected for you).

In case of any question send me an email.