| | |
Summary: Pattern Recognition Letters 14 (1993) 637 -645
North-Holland
PATREC IO77
August 1993
cedure previously proposed by the author (see
Aladjem (1991a,b)). The results obtained were con-
sistent for different data sets. This allows us to rate
the significance of the parameters, to select the ap-
propriate values and to find the guidelines for
parameter variations. The required variations are
reduced to a great extent which leads to a fast
search for a high quality mapping.
2. Nonparametric linear mapping procedure
In this section we present a short description of
the nonparametric linear mapping method proposed
by the author (Aladjem (1991a,b)). This method is
oriented to the discrimination of two classes e)1
and a)2. It is applied to data after previous nor-
malization
. x-Art, (l)
|
