Abstract | ||
---|---|---|
Investigating the pattern of host load in computing systems is very useful for discovering the data features and predicting the host load in the future. Since the host load can be regarded as the time series data, this paper proposes a pattern discovery framework for host load data by applying time series analysis methods. In the proposed framework, the effective data representation, data segmentation and feature extraction methods are designed based on the characteristics of the host load data. The DBSCAN clustering algorithm is then adopted in the pattern discovery framework to find the patterns in the host load. The extensive experiments have been conducted in this paper to verify the effectiveness of the proposed framework. |
Year | DOI | Venue |
---|---|---|
2014 | 10.1109/CSE.2014.78 | C3S2E |
Keywords | Field | DocType |
dbscan clustering algorithm,pattern discovery model,time series analysis,pattern clustering,data segmentation,data structures,data representation,feature extraction,data features,data mining,host load data,computing systems,time series data,time series,euclidean distance,clustering algorithms,vectors | Data mining,Time series,Data segment,External Data Representation,Computer science,Euclidean distance,Feature extraction,Artificial intelligence,Cluster analysis,Machine learning,DBSCAN,Computing systems | Conference |
Citations | PageRank | References |
2 | 0.37 | 21 |
Authors | ||
4 |
Name | Order | Citations | PageRank |
---|---|---|---|
Zhuoer Gu | 1 | 7 | 1.89 |
Cheng Chang | 2 | 12 | 4.35 |
Ligang He | 3 | 542 | 56.73 |
Kenli Li | 4 | 1389 | 124.28 |