Clustering techniques for mixed categorical and numeric datasets

Name and tittle: Amir Ahmad, Assistant Professor

Project Title: Clustering techniques for mixed categorical and numeric datasets

Abstract:

Many clustering algorithms have been proposed to cluster numeric datasets. However, most of these methods cannot handle mixed datasets consisting of categorical and numeric attributes. In this project, we will propose various transformation techniques to convert mixed datasets into pure numeric datasets with minimal information loss. Then, the clustering algorithms for numeric datasets will be applied to the new numeric datasets.

Kernels will also be used in these transformation techniques to use the representational power of kernels. Experiments will be carried out on publicly available datasets to show the effectiveness of the proposed approach.  

Share this news
 
Last updated:  Site best viewed at 1024 x 768 resolution in I.E 9+, Mozilla 3.5+, Google Chrome 3.0+, Safari 5.0+

عفوا

لايوجد محتوى عربي لهذه الصفحة

عفوا

يوجد مشكلة في الصفحة التي تحاول الوصول إليها

Feb 1, 2018