SOM++: integration of self-organizing map and k-means++ algorithms
MLDM'13 Proceedings of the 9th international conference on Machine Learning and Data Mining in Pattern Recognition
Hi-index | 0.00 |
This paper tries to find out a method of establishing a universal evaluation index system which is not only used to evaluate one single emergency plan, but also used to evaluate a class of emergency plans. According to the similarity of the content structures of emergency plans, statistical analysis is performed on their paragraph headings firstly, and then clustering is performed on the emergency plans using the method hybrid of SOM Network and K-means clustering algorithm, finally, we establish a universal evaluation index system for a cluster of emergency plans.