Shift and Scale Invariant Search of Multi-attribute Time Sequences

We investigate the problem of searching similarmulti-attribute time sequences in databases.Such sequences arise naturally in a number of medical,financial, video, weather forecast, and stock market databaseswhere more than one attribute is of interest at a timeinstant. We formulate a new symmetric scale and shift invariantnotion of distance for such sequences.We also propose a new indexstructure that transforms the data sequencesand clusters them according totheir shiftings and scalings. This clusteringimproves the efficiency considerably. According to ourexperiments with real and synthetic datasets, theindex structure\'s performance is 5 to 60 timesbetter than competing techniques, the exact speedup based on other optimizations such as caching and replication. Finally, we also consider the subsequence search problem.