The preselection and likelihood selection are optimised to separate the
and
events from the background
events.
As a result of the relative likelihood selection, events are classified as
or
events, or as both. For the ambiguous events,
a relative likelihood is constructed to discriminate between the two possible
cases, so all events can only be put in just one category.
However, approximately 30% of the events will actually be
events. So all events passing the
selection are reclassified as
or
events. A
similar procedure is undertaken for the events passing the
selection.
New relative likelihoods are calculated. The predominant
events
that are selected as
events are where the tau lepton decays into an
electron,
, or a one prong hadronic state,
.
Likelihoods are calculated for the event being from each of these two
processes using the same variables as before. Relative likelihoods are then
formed to discriminate between the
and
events. An
example of the relative likelihood
between
and
is shown below:
![]() |
(5.2) |
If any of these relative likelihoods are greater than 0.5 the event is
categorised as a
event. A similar procedure is applied to the
events to separate out the
and
events from
them.