Volume 7 Issue 2 - July 2016

  • 1. Effective trajectory data analysis using continuous k-means clustering

    Authors : Kanmani Palanisamy, Devendrakumar R.n, Ramalakshmi K

    Pages : 291-299

    DOI : http://dx.doi.org/10.21172/1.72.547

    Keywords : TKM(Threshold-based k-means monitoring method), HC*(an improved HC), confinement circle (CC),Continuous K-means clustering

    Abstract :

    Network monitoring tools are designed to analyze the network traffic status. Network transaction informations are observed from the frame headers. Data transmission, communication and computational overheads are analyzed from the network monitoring tools. Network load deviation estimation is a complex task. K-means clustering algorithm is used to estimate the network load. All the network transactions are passed into the clustering process. Threshold values are used to find out the load differences.The K-means clustering based monitoring model uses the centroid values to assess the network load difference. The system analyzes the network load under single server and multiple server environment. Network load adjustment process is initiated with reference to the load variations. Computational load and communication load are adjusted with the load variation information collected from the monitoring application. The system analyzes the node movement status.The proposed system is designed to improve the monitoring K-means based monitoring mechanism with distance measures. The distance measure is enhanced with support information. Support information reflects the attribute relationship with the entire transaction set. Attribute support functions are integrated with the distance estimation models. Dynamic distance measurement is used in the system.

    Citing this Journal Article :

    Kanmani Palanisamy, Devendrakumar R.n, Ramalakshmi K, "Effective trajectory data analysis using continuous k-means clustering", https://www.ijltet.org/journal_details.php?id=906&j_id=3251, Volume 7 Issue 2 - July 2016, 291-299, #ijltetorg