Comparison of Two Different Prediction Schemes for the Analysis of Time Series of Graphs

This paper is concerned with time series of graphs and compares two novel schemes that are able to predict the presence or absence of nodes in a graph. Our work is motivated by applications in computer network monitoring. However, the proposed prediction methods are generic and can be used in other applications as well. Experimental results with graphs derived from real computer networks indicate that a correct prediction rate of up to 97% can be achieved.

[1]  Horst Bunke,et al.  Graph-Based Tools for Data Mining and Machine Learning , 2003, MLDM.

[2]  Rainer Schmidt,et al.  Temporal Abstractions and Case-Based Reasoning for Medical Course Data: Two Prognostic Applications , 2001, MLDM.

[3]  Horst Bunke,et al.  Matching graphs with unique node labels , 2004, Pattern Analysis and Applications.

[4]  Kuniaki Uehara,et al.  Discover Motifs in Multi-dimensional Time-Series Using the Principal Component Analysis and the MDL Principle , 2003, MLDM.

[5]  Ambuj K. Singh,et al.  Optimizing similarity search for arbitrary length time series queries , 2004, IEEE Transactions on Knowledge and Data Engineering.

[6]  Wai Lam,et al.  News Sensitive Stock Trend Prediction , 2002, PAKDD.

[7]  Lior Rokach,et al.  CHANGE DETECTION IN CLASSIFICATION MODELS INDUCED FROM TIME SERIES DATA , 2004 .

[8]  J. Ross Quinlan,et al.  C4.5: Programs for Machine Learning , 1992 .

[9]  Richard J. Povinelli,et al.  A New Temporal Pattern Identification Method for Characterization and Prediction of Complex Time Series Events , 2003, IEEE Trans. Knowl. Data Eng..

[10]  Eamonn J. Keogh,et al.  Segmenting Time Series: A Survey and Novel Approach , 2002 .

[11]  Abraham Kandel,et al.  Data Mining in Time Series Database , 2004 .

[12]  Philip S. Yu,et al.  Mining Asynchronous Periodic Patterns in Time Series Data , 2003, IEEE Trans. Knowl. Data Eng..

[13]  Horst Bunke,et al.  Detection of Abnormal Change in a Time Series of Graphs , 2002, J. Interconnect. Networks.

[14]  Dimitrios Gunopulos,et al.  INDEXING TIME-SERIES UNDER CONDITIONS OF NOISE , 2004 .