A Search for the SM Higgs Boson by Burt DeWilde Kalamazoo College Against all odds.

A Search for theSM Higgs Boson

by Burt DeWilde

Kalamazoo College

Against all odds

The Standard Model (again)

Particles grouped into generations by mass

Higgs field with non-zero VEV causes EWSB -> mass

Higgs boson is the particle associated with this field

The Tevatron

The D0 Detector

Run I vs. Run II

Increase in center-of-mass energy->1.96TeV

Increase in number of bunches, decrease in spacing of bunches

Much higher luminosity (Hopefully) enough for

Higgs boson detection

Higgs Production Modes

Higgs Decay Modes

bb channel dominant up to M=135GeV

Z boson decays into two muons less frequent, but also cleaner channel

Neural Network Structure

NN Variables: Signal, Background, and Data

NN Variables: Signal, Background, and Data

Training Method Comparison

Signal Significancevs. Training Method

Optimal # of Epochsfor # of Hidden Neurons

Signal Significance vs.Optimal # of Neurons/Epochs

Signal Significance vs.NN Weighting Scheme

Invariant Mass Distributionsfor Potential Higgs Masses

Signal Significance vs.Invariant Mass Base Cut

NN Results forHiggs M=105GeV

NN Results forHiggs M=115GeV

NN Results forHiggs M=125GeV

NN Results forHiggs M=135GeV

NN Results forHiggs M=145GeV

NN Results forHiggs M=155GeV

Importance of the Results

NN achieves better signal/background discrimination than cuts on Mjj alone

NN analysis gives better signal significance than conventional analysis

Integrated luminosity required for Higgs detection reduced 40-70%

Not too shabby

Acknowledgements

Andy Haas, for meaning, direction, and data The Nevis crew, for troubleshooting and

much-needed moral support The National Science Foundation, for

research experience and money