Escolar Documentos
Profissional Documentos
Cultura Documentos
Available at http://www.ijcsonline.com/
Abstract
In this paper we considered data bases with attributes whose values are categorical. These values cannot be ordered in
the single way and clustering of data is challenge. Categorical data have no single ordering. It can be viewed based on
the specific ordering. We need different methods and different similarity measures to discover natural grouping of
categorical data. Hence we use several clustering algorithms like k-mods algorithm, ROCK algorithm, STIRR algorithm,
CACTUS algorithm to summarize the characteristics of categorical data.
Keywords: Clustering Algorithms, Data mining.
I.
INTRODUCTION
their
where
Where
and
are the numbers of objects in the
data set with attributes values
and for the attribute i.
the mode of the set is the values that appears the most in
244 | International Journal of Computer Systems, ISSN-(2394-1065), Vol. 03, Issue 03, March, 2016
Sajikumar S. et al
Data
Label data on
Overview
Disk of ROCK
III.
and
IV.
245 | International Journal of Computer Systems, ISSN-(2394-1065), Vol. 03, Issue 03, March, 2016
Sajikumar S. et al
V.
2. additional operator
3. the generalization of the addition operator that is
called the Sp combining rule , where p is an odd
natural number. Sp(w1..,Wk=w1p+.+wk p)(1/p).
Addition is simply an S1 rule.
4. A limiting version of the Sp rules, which is referred
to as { w1..wk } is equal to wi, where wi is the
largest absolute value among the weights in {
w1..wk }
246 | International Journal of Computer Systems, ISSN-(2394-1065), Vol. 03, Issue 03, March, 2016
Sajikumar S. et al
VI.
And a cluster is defined by the following
Cluster in CACTUS :
C i ..C n is a cluster if and only if C i is maximal
for all i and Support(c) is times the expected. The above
definition implies that clusters could be region, where
region {a1,a2}{b1,b2}{c1,c2},(dotted area),defines the
cluster after that can delve into the stages of
CACTUS. In the summarization phase, two types of
summaries are computed:
CONCLUSION
REFERENCES
[1]
2. Intra-attribute
summaries:
computation
of
similarities between attribute values of the same
attribute.
[2]
[3]
[4]
[5]
[6]
[7]
[8]
[9]
247 | International Journal of Computer Systems, ISSN-(2394-1065), Vol. 03, Issue 03, March, 2016