HEPP Workshop 2019

Applications of Machine Learning techniques in mono-higgs searches

1 Feb 2019, 12:30

20m

Room 0017 (Nelson Mandela University)

Session V

Speaker

Mr Moratwe Mdluli (University of the Witwatersrand (ZA))

Description

A search for the Dark Matter (DM) in association with the Standard Model like Higgs boson was performed during the 2015 and 2016 data. In this analysis we use Multivariate Analysis (MVA) methods to suppress the fake missing transverse energy, using the Boosted Decision Tree (BDT) we are able to enhance the regression performance of typically weak MVA methods for the DM search. We scan the results on the cut of MET and METSig to find the best MVA training and how much we can gain and which variable is the best.

Presentation materials

HEPP2019 Presentation.pdf