Title | ||
---|---|---|
On the applicability of mobility metrics for user movement pattern recognition in MANETs |
Abstract | ||
---|---|---|
In this paper we propose a set of mobility metrics, which are employed in the generation of supervised classification learning methods through the decision tree algorithm, with the goal to recognize user movement patterns in mobile ad hoc networks. Hundreds of scenarios produced by several well-known mobility models were employed for training and testing the supervised algorithms. The most suitable classification model showed an accuracy of 99.20% and Kappa index of 0.991, which indicates a high level of agreement between the classification model and real classification. |
Year | DOI | Venue |
---|---|---|
2013 | 10.1145/2508222.2508228 | MOBIWAC |
Keywords | Field | DocType |
decision tree algorithm,supervised algorithm,supervised classification,user movement pattern recognition,classification model,well-known mobility model,mobility metrics,suitable classification model,kappa index,real classification,high level,pattern recognition,mobility model | Mobile ad hoc network,Data mining,One-class classification,Pattern recognition,Computer science,Mobility model,Artificial intelligence,Machine learning,Decision tree learning | Conference |
Citations | PageRank | References |
1 | 0.35 | 13 |
Authors | ||
2 |
Name | Order | Citations | PageRank |
---|---|---|---|
Elmano R. Cavalcanti | 1 | 1 | 0.35 |
Marco Aurélio Spohn | 2 | 130 | 12.12 |