The Art of Differentiating Computer Programs - Uwe Naumann

Résumé :
The first entry-level book on algorithmic (also known as automatic) differentiation (AD), providing fundamental rules for the generation of first- and higher-order tangent-linear and adjoint code. The author covers the mathematical underpinnings as well as how to apply these observations to real-world numerical simulation programs.

Biographie:
Uwe Naumann is Associate Professor for Computer Science at RWTH Aachen University. He has co-chaired international conferences on AD held in Nice (2000), Chicago (2004) and Bonn (2008) and co-edited the proceedings for those conferences. His professional interests are primarily motivated by AD. They range from compilers for the semantic transformation of numerical programs via corresponding (domain-specific) program analyses and combinatorial algorithms to high-performance scientific computing and the application of adjoint methods in computational science, engineering and finance. He has co-authored more than 80 publications and is a member of SIAM and of the Numerical Algorithms Group, Ltd.

Sommaire:
Preface; Acknowledgements; Optimality; 1. Motivation and introduction; 2. First derivative code; 3. Higher derivative code; 4. Derivative code compilers - an introductory tutorial; 5. dcc - a prototype derivative code compiler; Appendix A. Derivative code by overloading; Appendix B. Syntax of dcc input; Appendix C. (Hints on) Solutions; Bibliography; Index.