Abstract
In this paper, the problem of detecting the major changes in the stream of service requests is formulated. The change of stream component varies over time and depends on, e.g., a time of a day. The underlying cause of the change is called a context. Hence, at each moment there exists a probability distribution determining the probability of requesting the system service conditioned by the context. The aim is to find such a moment in which the distributions change. To solve that problem dissimilarity measures between two probability distributions are given. Nevertheless, detecting every change is not interesting but only long-lasting changes because of the costs of the service system resources reallocation. Therefore, in the proposed algorithm an issue of sensitivity to temporary changes detection is considered.
Original language | English |
---|---|
Title of host publication | Knowledge-Based and Intelligent Information and Engineering Systems - 15th International Conference, KES 2011, Proceedings |
Pages | 591-600 |
Number of pages | 10 |
Edition | PART 2 |
DOIs | |
Publication status | Published - 29 Sept 2011 |
Externally published | Yes |
Event | 15th International Conference on Knowledge-Based and Intelligent Information and Engineering Systems, KES 2011 - Kaiserslautern, Germany Duration: 12 Sept 2011 → 14 Sept 2011 |
Publication series
Name | Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics) |
---|---|
Number | PART 2 |
Volume | 6882 LNAI |
ISSN (Print) | 0302-9743 |
ISSN (Electronic) | 1611-3349 |
Conference
Conference | 15th International Conference on Knowledge-Based and Intelligent Information and Engineering Systems, KES 2011 |
---|---|
Country/Territory | Germany |
City | Kaiserslautern |
Period | 12/09/11 → 14/09/11 |
Funding
Acknowledgments. The research presented in this paper has been partially supported by the European Union within the European Regional Development Fund program no. POIG.01.03.01-00-008/08. The research has been also partially co-financed by European Union within European Social Fund.
Keywords
- datastream
- dissimilarity measure
- sliding window estimation