CLASSIF1 Data Pattern Classification
fig.16: Reclassification of the Learning Set
of Random Number Data
- The category classification mask (catmask#1) for category#1 records
(cat#1) as reference group contains a sequence of (0)
characters because the majority (70%) of the parameter values are
located between the two percentile thresholds 15% and 85%,
with 15% of the parameter values below (-) the lower percentile
and 15% above (+) the upper percentile.
- The automatically selected discriminatory parameters (catmask#2) for
category#2 records (cat#2) are all increased (+).
- Data records are classified according to their highest positional
coincidence with either one of the two category classification masks
catmask#1 and catmask#2.
- The degree of coincidence with the best fitting category mask
is expressed as mask coincidence factor.
The factor is 1.00 for data records #003,013,025,006,008,024,026
containing (0) or (-) but no (+) triple matrix characters.
- Displayed are the first 10 data patterns of cat#1 and cat#2 data records.
Records #001,002,009,010,019,020,029,030,039,040 are invisible.
They were removed prior to the learning phase and serve as
unknown test set.
This permits to independently assess the discriminatory potential
of learned classifiers for unknown data records.
Last Update: Feb 01,2020
First display: Feb 16,2005