Influence angel cluster approach for data clustering
Clustering allows one to handle a large data set effectively. It is a technique for solving classification problems. There are two major challenges in clustering. First, identifying clusters in high-dimensional data sets is a difficult task because of the curse of dimensionality. Second, a new dissi...
| Main Authors: | , |
|---|---|
| Format: | Conference or Workshop Item |
| Language: | English |
| Published: |
2010
|
| Subjects: | |
| Online Access: | https://repo.uum.edu.my/id/eprint/2208/1/Nazrina_Aziz_%26_Dong_Qian_Wang.pdf |
| _version_ | 1825739933917118464 |
|---|---|
| author | Aziz, Nazrina Dong, Qian Wang |
| author_facet | Aziz, Nazrina Dong, Qian Wang |
| author_sort | Aziz, Nazrina |
| collection | UUM |
| description | Clustering allows one to handle a large data set effectively. It is a technique for solving classification problems. There are two major challenges in clustering. First, identifying clusters in high-dimensional data sets is a difficult task because of the curse of dimensionality. Second, a new dissimilarity measures is needed as some traditional distance functions cannot capture the pattern dissimilarity among the objects. This article interested in the latter challenge. Notice that data measures are very important steps in clustering. There are many types of data measurement that deal with continuous,categorical or mixed variables. This article proposed an alternative measurement, called Influence Angle Cluster Approach (iaca). The iaca was developed based on eigenstructure
of the covariance matrix. The proposed measurement able to identify cluster of observation and it also has the ability to handle a data set with mixed variables. Apart from developing a cluster for a data set, this study also measure whether or not the proposed IACA have constructed either a strong or reasonable clustering structure by using silhouette index. |
| first_indexed | 2024-07-04T05:18:24Z |
| format | Conference or Workshop Item |
| id | uum-2208 |
| institution | Universiti Utara Malaysia |
| language | English |
| last_indexed | 2024-07-04T05:18:24Z |
| publishDate | 2010 |
| record_format | eprints |
| spelling | uum-22082011-02-20T08:50:15Z https://repo.uum.edu.my/id/eprint/2208/ Influence angel cluster approach for data clustering Aziz, Nazrina Dong, Qian Wang QA Mathematics Clustering allows one to handle a large data set effectively. It is a technique for solving classification problems. There are two major challenges in clustering. First, identifying clusters in high-dimensional data sets is a difficult task because of the curse of dimensionality. Second, a new dissimilarity measures is needed as some traditional distance functions cannot capture the pattern dissimilarity among the objects. This article interested in the latter challenge. Notice that data measures are very important steps in clustering. There are many types of data measurement that deal with continuous,categorical or mixed variables. This article proposed an alternative measurement, called Influence Angle Cluster Approach (iaca). The iaca was developed based on eigenstructure of the covariance matrix. The proposed measurement able to identify cluster of observation and it also has the ability to handle a data set with mixed variables. Apart from developing a cluster for a data set, this study also measure whether or not the proposed IACA have constructed either a strong or reasonable clustering structure by using silhouette index. 2010 Conference or Workshop Item NonPeerReviewed application/pdf en https://repo.uum.edu.my/id/eprint/2208/1/Nazrina_Aziz_%26_Dong_Qian_Wang.pdf Aziz, Nazrina and Dong, Qian Wang (2010) Influence angel cluster approach for data clustering. In: 2nd International Conference on Mathematical Sciences (ICMS2 2010), 30 November - 3 December 2010 , Putra World Trade Centre (PWTC) Kuala Lumpur, Malaysia . (Unpublished) http://pkukmweb.ukm.my/~ppsmfst/icms2/ |
| spellingShingle | QA Mathematics Aziz, Nazrina Dong, Qian Wang Influence angel cluster approach for data clustering |
| title | Influence angel cluster approach for data clustering |
| title_full | Influence angel cluster approach for data clustering |
| title_fullStr | Influence angel cluster approach for data clustering |
| title_full_unstemmed | Influence angel cluster approach for data clustering |
| title_short | Influence angel cluster approach for data clustering |
| title_sort | influence angel cluster approach for data clustering |
| topic | QA Mathematics |
| url | https://repo.uum.edu.my/id/eprint/2208/1/Nazrina_Aziz_%26_Dong_Qian_Wang.pdf |
| work_keys_str_mv | AT aziznazrina influenceangelclusterapproachfordataclustering AT dongqianwang influenceangelclusterapproachfordataclustering |