Prediction of Occurrence of Discrete Events - Geomathematics: Theoretical Foundations, Applications and Future Developments - page 152

Fig. 5.10 Test of overall
conditional independence,
using the Kolmogorov-
Smirnov statistic. Note that
the observed curve ( open
circles ) stays within the
95 % confidence envelope
surrounding the predicted
curve ( solid line ) (Source:
Bonham-Carter et al. 1990 ,
Fig. 3)

1.0

0.8

0.6

0.4

0.2

0.0
0.0 0.1 0.2 0.3 0.4

Fig. 5.11 Posterior
probability plotted against
cumulative area, with
producing gold mines
shown as circles whose
radii reflect magnitudes
of reported production
(Source: Bonham-Carter
et al. 1990 , Fig. 4)

0.4

SALMON RIVER, 41.6

0.3

GOLDENVILLE, 210.2
UPPER SEAL HARBOUR, 57.8
COCHCRANE HILL, 2.1
ISAACS HARBOUR, 39.7
LOWER SEAL HARBOUR, 34.3
HARIGAN COVE, 7.9
WINE HARBOUR, 42.7
FOREST HILL, 25.1
COUNTRY HARBOUR, 9.9
ECUM SECUM, 1.3

0.2

0.1

0.0
100 80 60

Cumulative Area, %

40 20 0

drift layer. The posterior probability map shown in Fig. 5.17 is based on buffer zone
around ORM and steep slope zone only. It not only quantifies the relationship between
the known artesian aquifers and these two binarymap layers, it also outlines areas with
no or relatively few aquifers that have good potential for additional aquifers.

5.1.4 Variance of the Contrast and Incorporation
of Missing Data

For estimation of the variances of weights and contrasts in Weights-of-Evidence,
use was made of asymptotic likelihood expressions ( cf . Bishop et al. 1975 ,