Using the small-angle approximation to calculate the Higgs mass in Hbb events. Andy has made plots of the approximate mass for different dR values.
The smaller dR values look like Poisson distributions - why might this be? What kind of distribution are we expecting?
What kind of things give us Poisson distributions?
Getting the actual mass with mCalc gives a nice Gaussian Higgs mass in Hbb events, and what looks like an exponentially decaying distribution with a cliff-edge around m_bb=80 GeV for the background.
Why might there be a cliff edge at m_bb=80? Does this change if we alter our selection cuts?
Would you expect the background to be falling off exponentially?
In the classification using px versus py, the background events are heavily subsumed by the signal.
Does this mean that you do not have enough background events passing the cuts? How will you address this?
Putting a lower cut of 40 GeV on the bs will not make the pT distributions of the signal and background the same. The signal will be harder and the background softer, as you are seeing on your classifier plots. Is there some way that you can address this, for example by applying an upper cut in addition, or in binning your events in e.g, pT(average) b, with bins like 40-50, 50-60, ... ?
Notes on Andy's email update:
Using the small-angle approximation to calculate the Higgs mass in Hbb events. Andy has made plots of the approximate mass for different dR values.
Getting the actual mass with mCalc gives a nice Gaussian Higgs mass in Hbb events, and what looks like an exponentially decaying distribution with a cliff-edge around m_bb=80 GeV for the background.
In the classification using px versus py, the background events are heavily subsumed by the signal.