[Colloq] PhD Thesis Proposal, Huanmei Wu, March 17

[Rachel Kalweit]

College of Computer and Information Science
PhD Thesis Proposal:  Structured Time Series Stream Data


SPEAKER:  Huanmei Wu
TIME:   Thursday, March 17, 2:00pm
PLACE:  366 West Village H
Advisor:   Betty Salzberg


Management of time series stream data is an important technique for 
database, networks, operating systems, and theoretical research. 
Analysis of time series stream data is widely used for many applications 
such as economic forecasting, stock market analysis, budgetary analysis, 
nd workload projects.

We have developed a framework to achieve these objectives. The framework 
composes (a) a piecewise linear model and finite state automaton which 
captures the internal structure of a time series stream data, (b) an 
online segmentation and pruning algorithm which produces the piecewise 
linear representation of raw stream data in real-time, (c) an online 
subsequence matching engine which generates dynamic query subsequences, 
defines application specific subsequence similarity measures and 
performs similarity matching with consideration of the internal 
structure of the time series and (d) statistical and probabilistic 
techniques which analyze similarity matching results for many specific 
applications, such as motion characterization, future prediction, 
correlation discovery and clustering.

Our approach covers all application domains with structured time series 
stream data. Our techniques can be used for real-time systems where 
response time is critical. We have applied the framework to multiple 
problem domains, such as financial data analysis and tumor respiratory 
motion characterization (used in cancer radiation treatment).