A Regression-Based Temporal Pattern Mining Scheme for Data Streams Wei-Guang Teng, Ming-Syan Chen, and Philip S. Yu   View Paper (PDF)   Return to Data Mining/Streams (Session A2) Abstract We devise in this paper a regression-based al- gorithm, called algorithm FTP-DS (Frequent Temporal Patterns of Data Streams), to mine frequent temporal patterns for data streams. While providing a general framework of pat- tern frequency counting, algorithm FTP-DS has two major features, namely one data scan for online statistics collection and regression- based compact pattern representation.To at- tain the feature of one data scan, the data segmentation and the pattern growth scenar- ios are explored for the frequency counting purpose. Algorithm FTP-DS scans online transaction flows and generates candidate fre- quent patterns in real time. The second im- portant feature of algorithm FTP-DS is on the regression-based compact pattern repre- sentation. Specifically, to meet the space constraint, we devise for pattern representa- tion a compact ATF (standing for Accumu- lated Time and Frequency) form to aggre- gately comprise all the information required for regression analysis. In addition, we de- velop the techniques of the segmentation tun- ing and segment relaxation to enhance the functions of FTP-DS. With these features, al- gorithm FTP-DS is able to not only conduct mining with variable time intervals but also perform trend detection effectively. Synthetic data and a real dataset which contains net- work alarm logs from a major telecommunica- tion company are utilized to verify the feasi- bility of algorithm FTP-DS. ©2004 Association for Computing Machinery