Advanced numerical method is a course to introduce and to survey the
<br>recent development of numerical algorithms.  The topics may include
<br>
<br>1. Numerical algorithms on emerging computational architectures, such
<br>as GPU or multicore.
<br>2. Parallel numerical algorithms.
<br>3. Numerical algorithms for massive data set.
<br>4. Numerical methods for solving large linear systems and
<br>pre-conditioning techniques.
<br>5. Numerical methods for solving large eigenvalue problems.
<br>6. Numerical methods for matrix functions.
<br>7. Numerical methods for optimization.
<br>8. Numerical methods for tensor and high dimensional data.
<br>9. Fourier based method
<br>10. Sparse matrix technique
<br>
<br>For each topic, we will focus on its algorithms and numerical properties,
<br>accuracy, performance, and data storage.
<br>Students who take this course need to attend the lecture, present
<br>papers, and implement some of the algorithms.
<br>Implementation may use Matlab, C, or Fortran.
<br>
<br>Pre-request: linear algebra, C programming, (parallel programming is
<br>preferred, but not required.)