Segmentation of Continuous Data Streams Based on a Change Detection Methodology - Advanced Techniques in Knowledge Discovery and Data Mining - page 114

Table 4.3. Results of the XP hypothesis testing on the “Dropout” database
(1- P value).
Year

Reg
ion

Class Divisions Sump A vep Tr_Per Dropout

1997 98.6% 100% 67.1% 100.0% 99.6% 96.6% 99.8%

1998 99.8% 76.7% 88.6% 99.6% 51.1% 100% 100%

1999 99.1% 99.8% 99.6% 100% 97.7% 99.9% 100%

2000 93.1% 98.9% 100% 100% 96.4% 65.9% 100%

Table 4.4. Outcomes of implementing the change detection methodology on “Dropout”
database (1- P value).

Year CD XP(target)

1996 - -

1997 100.0% 99.8%

1998 100.0% 100.0%

1999 82.2% 100.0%

2000 100.0% 100.0%

105.0%

100.0% 100.0% 100.0%

100.0%

100.0% 100.0% 100.0%

99.8%

95.0%

90.0%

87 .7 %

85.0%

80.0%

1997 1998 1999 2000

Year

CD XP

Fig. 4.1. Summary of implementing the change detection methodology on
“Dropout” database (1 - P value).