Information Technology Reference
In-Depth Information
Degree Distribution
Degree Distribution
Degree Distribution
1
5
50
500
1
2
5
10
20
1
2
5
10
20
degree
degree
degree
Degree Distribution
Degree Distribution
Degree Distribution
12
5 0 0 0
1
2
5
10
50
200
12
5 0 0 0
degree
degree
degree
Degree Distribution
Degree Distribution
12 5 0 0 0
1
5
50
500
degree
degree
Fig. 3
Cumulative degree distributions of the eight ego networks
To measure how close our estimated distributions are to the real distribution, we
need to define a distance (difference) measure that quantifies to what extent the
estimated results are similar to the truth. Of course, the selection of a distance is
crucial for the outcome of a study. Many distance measures have been defined in the
literature. We use both Kolmogorov-Smirnov (KS)
D
statistic and Kullback-Leibler
Divergence. The KS
D
statistic is based on the maximum distance between the two
