This research project aimed to use an open source, feed forward neural network interfaced with ROOT to improve the calculation of the Bs transverse momentum. If successful, this method would be integrated into the DØ code structure.
| Date | Spring 2006 |
|---|---|
| Primary Goals | To implement a neural network to improve the calculation of the Bs transverse momentum from its daughter particles. |
| Software | root, nnf, emacs |
| Hardware | N/A |
Bs--> v_e + Ds; Ds --> mu+ + mu-
The Bs has too short of a lifetime to be detected in the DØ detector, so
its properties must be measured indirectly via its daughter particles. The
catch is that one of its daughter particles happens to be a neutrino,
resulting in energy and momentum information loss. The standard DØ coping
method is to scale the ptDs by a constant 'K factor' of 0.8595.
In the end it was found that the neural net could not improve upon the ptBs calculation.
Example Source Code
STARS Symposium Poster
