work

1
Not a
Bayesian analysis at
work
War
troublesome examples
J. Charles, A. Hoecker, H. Lacker, F. Le
Diberder, S. tJampens
2
Outline

• Setting the stage
• Bayesian analysis
• Simplified stage
• Concerns and Conclusion

3
Setting the Stage
B decays into p p final states
6 observables And 6 theoretical unknowns
6 decay amplitudes are involved
CP Violation
a
and Mirror Solutions
4

Bayesian Analysis
Data
a

• Theoretical parameters
• are treated as random variables
• - following known PDF

mind
5
Quantum Mechanics parameterization
6 decay amplitudes are involved
One phase may be set to zero
Bayesian (I)
6
Standard Model parameterization
The 6 amplitudes are split into two kind of
contributions
Trees Penguins
amplitudes phases
Bayesian (II)
Minuit output
7
Changing decay channel (??)
Bayesian (I)
Bayesian (II)
Bayesian (III)
amplitudes phases
Quantum Mechanics
real imaginary parts
8
B ? p p
Taking key experimental information away
removing all experimental hints
related to a
amplitudes phases
9
Spheres within a cube
Simplified Stage
6 D space
Bayesian treatment
one knows nothing about the individual space
components x,y,z
What do we known about the radius r
v(x2y2) ?
From nothing we deduce the radius
10
Conclusion
that attractive ?
priors
parameterization