Record number :
2346067
Title of article :
Constraint-based clustering and its applications in construction management
Author/Authors :
Cheng، نويسنده , , Ying-Mei and Leu، نويسنده , , Sou-Sen، نويسنده ,
Issue Information :
روزنامه با شماره پیاپی سال 2009
Pages :
7
From page :
5761
To page :
5767
Abstract :
Both mixed data types and cluster constraints are frequently encountered in the classification problems of construction management. For example, in a bridge let project, engineers generally group the bridges into several subgroups based on their proximities, structure type, material, etc. Moreover, constraints may be set for each cluster to ensure the project’s overall effectiveness. In this study, an effective clustering algorithm – the constrained k-prototypes (CKP) algorithm – is proposed to resolve the abovementioned problems. Several tests and experimental results have shown that CKP cannot only handle mixed data types but also satisfy user-specified constraints. In order to demonstrate the applicability of CKP, it is also applied to real-world problems in construction management.
Keywords :
Affinity diagram , Constraint-based clustering , construction management , K-Means , k-Prototypes
Journal title :
Expert Systems with Applications
Journal title :
Expert Systems with Applications
Serial Year :
2009
Link To Document :
بازگشت