recent development of numerical algorithms. The topics may include

1. Review of linear algebra.

2. Sparse matrix techniques. (representation, multiplication, direct solvers)

3. Recursive algorithms. (FFT, Fast Multipole method, Multigrid)

4. Communication complexity.

5. Mixed precision algorithms.

6. Matrix functions.

7. Numerical methods for tensor and high dimensional data.

For each topic, we will focus on its algorithms and numerical properties,

accuracy, performance, and data storage. Students who take this course

need to attend the lecture, present papers, and implement some of the

algorithms. Implementation may use Matlab, C, or Fortran.

Pre-request: linear algebra, programming