We present DSA - Derivative time series Segment Approximation, a novel representation model for time series designed for effective and efficient similarity search. DSA substantially exploits derivative estimation, segmentation and dimensionality reduction to meet at least the requirements of high sensitivity to main features (trends) of time series and robustness to outliers. Experiments show that DSA is drastically faster and still as good or better than the prominent state-of-the-art similarity methods.
Effective and Efficient Similarity Search in Time Series
GRECO, Sergio;TAGARELLI, Andrea
2006-01-01
Abstract
We present DSA - Derivative time series Segment Approximation, a novel representation model for time series designed for effective and efficient similarity search. DSA substantially exploits derivative estimation, segmentation and dimensionality reduction to meet at least the requirements of high sensitivity to main features (trends) of time series and robustness to outliers. Experiments show that DSA is drastically faster and still as good or better than the prominent state-of-the-art similarity methods.File in questo prodotto:
Non ci sono file associati a questo prodotto.
I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.