Laurent Locatelli

1
Background discrimination using a Neural Network
Laurent Locatelli Laurent.Locatelli_at_epfl.ch LHCb
CERN Jets Working Group 26th April 2006
2
Pattern(1) Mj1 Pattern(2) Ej1
Pattern(3) Mj2 Pattern(4) Ej2
Pattern(5) Ptj1 Pattern(6) Ptj2
Pattern(7) etaj1 Pattern(8) etaj2
Pattern(9) Pplep1 Pattern(10) Ptlep1
Pattern(11) Elep1 Pattern(12) Pt(j1
j2 lep1) Pattern(13) Pttot
Pattern(14) Mtot Pattern(15) Etot
Pattern(16) M(tot - lep1) Pattern(17)
dR(j1 - lep1) Pattern(18) dR(j2 - lep1)
3
(No Transcript)
4
(No Transcript)
5
(No Transcript)
6
(No Transcript)
7
(No Transcript)
8
(No Transcript)
9
(No Transcript)
10
(No Transcript)
11
(No Transcript)
12
(No Transcript)
13
(No Transcript)
14
(No Transcript)
15
Next step Combine the results of the separate
Neural Networks to
discriminate the background from the signal (HA).
Out
1
Out
2
Out
3
If the unknown event is a signal (HA), the 6
Neural Networks should give an answer close to
1. Therefore the product of the outputs OUT
should also be close to 1. A cut applied on the
variable OUT has to be studied.
Out
4
Out
5
Out
6
16
(No Transcript)
17
(No Transcript)
18
(No Transcript)
19
(No Transcript)
20
(No Transcript)
21
(No Transcript)
22
(No Transcript)