Implementation and Test of Numerical Optimization Algorithms in C/C++ | Christian Wichmann Moesgaard
| | Abstract | The goal of this report is to describe the twice continuously differentiable unconstrained optimization problem. Recognizing that it is impossible to check every point, it provides the mathematical foundation for the BFGS algorithm, to solve the problem locally.
Once this is introduced, the report moves on to a practical implementation - the steps required to begin using scientific computing by including BLAS and LAPACK. It then provides the documentation for the implementation in C and an example of how to use it.
Then, it describes how the optimization algorithms can be called from C, C++, Objective-C++, and FORTRAN.
Finally, it does some performance tests, and it is found that libopti is fast compared to scripting based solutions, and as much as 7-8 times faster than the GNU Scientific Library, however it is also not entirely reliable for solving every problem in its class. In fact, no implementation is tested which could solve every problem in its class. | | Type | Bachelor thesis [Academic thesis] | | Year | 2012 | | Publisher | Technical University of Denmark, DTU Informatics, E-mail: reception@imm.dtu.dk | | Address | Asmussens Alle, Building 305, DK-2800 Kgs. Lyngby, Denmark | | Series | IMM-B.Sc.-2010-36 | | Note | | | Electronic version(s) | [pdf] | | Publication link | http://www.imm.dtu.dk/English.aspx | | BibTeX data | [bibtex] | | IMM Group(s) | Scientific Computing |
|