A subsequence matching algorithm supporting moving average transform of arbitrary order in time-series databases using index interpolation

Proposes a subsequence matching algorithm that supports a moving-average transform of arbitrary order in time-series databases. The existing subsequence matching algorithm of C. Faloutsos et al. (1994) requires an index for each moving-average order, which causes serious storage and CPU time overheads. In this paper, we solve the problem using index interpolation. The proposed algorithm can use only a few indexes for pre-selected moving-average orders k, and it performs subsequence matching for an arbitrary order m (⩽k). We prove that the proposed algorithm causes no false dismissal. For selectivities less than 10^-2, the degradation of the search performance compared with the fully-indexed case is no more than 17.2% when two out of 128 indexes are used. The algorithm works better with smaller selectivities