Clustering hoax fire calls using evolutionary computation technology

Authors:
Lili Yang;Michael Gell;Christian W. Dawson;Martin R. Brown
Affiliations:
School of Computing and Technology, University of Derby, UK;School of Computing and Technology, University of Derby, UK;Computer Science Department, Loughborough University, Loughborough, UK;Department of Computing, University of Central Lancashire, Lancashire, UK
Venue:
IEA/AIE'2003 Proceedings of the 16th international conference on Developments in applied artificial intelligence
Year:
2003

Citing 3
Cited 0

Data mining

Data mining
A Faster Genetic Clustering Algorithm

Real-World Applications of Evolutionary Computing, EvoWorkshops 2000: EvoIASP, EvoSCONDI, EvoTel, EvoSTIM, EvoROB, and EvoFlight
Clustering with a genetically optimized approach

IEEE Transactions on Evolutionary Computation

Quantified Score

Hi-index	0.00

Visualization

Abstract

Hoax fire calls put an unnecessary burden on service resources and endanger life by making personnel and appliances unavailable for genuine incidents. Identifying the higher risk areas of hoax fire calls will be helpful in reducing the hoax calls In this paper, the hoax caller is located by a 6-figure map reference with two lead letters. A GA based evolutionary computation technology is proposed and applied to cluster the hoax calls into several groups according to their locations. The number of clusters is fixed at each GA run, and it is incremented by 1 for each iteration until the desired fitness (quality of the clustering partition) is achieved. The novel fitness function allows each cluster geographically covering a similar size of the areas and avoids empty clusters occur. The algorithm is then applied to the identification of higher risk areas of hoax fire calls. A spatial visualization is also used to display the clustering results in which three higher risk areas are clearly identified.