Efficient Similarity Search In Sequence Databases
FODO '93 Proceedings of the 4th International Conference on Foundations of Data Organization and Algorithms
VLDB '95 Proceedings of the 21th International Conference on Very Large Data Bases
A symbolic representation of time series, with implications for streaming algorithms
DMKD '03 Proceedings of the 8th ACM SIGMOD workshop on Research issues in data mining and knowledge discovery
New Time Series Data Representation ESAX for Financial Applications
ICDEW '06 Proceedings of the 22nd International Conference on Data Engineering Workshops
A decade of progress in indexing and mining large time series databases
VLDB '06 Proceedings of the 32nd international conference on Very large data bases
Experiencing SAX: a novel symbolic representation of time series
Data Mining and Knowledge Discovery
ISITC '07 Proceedings of the 2007 International Symposium on Information Technology Convergence
Representing financial time series based on data point importance
Engineering Applications of Artificial Intelligence
iSAX: indexing and mining terabyte sized time series
Proceedings of the 14th ACM SIGKDD international conference on Knowledge discovery and data mining
Two Novel Adaptive Symbolic Representations for Similarity Search in Time Series Databases
APWEB '10 Proceedings of the 2010 12th International Asia-Pacific Web Conference
Hi-index | 0.00 |
Existing symbolic approaches for time series suffer from the flaw of missing important trend feature, especially in financial area. To solve this problem, we present Trend-based Symbolic approximation (TSX), based on Symbolic Aggregate approximation (SAX). First, utilize Piecewise Aggregate Approximation (PAA) approach to reduce dimensionality and discretize the mean value of each segment by SAX. Second, extract trend feature in each segment by recognizing key points. Then, design multiresolution symbolic mapping rules to discretize trend information into symbols. Experimental results show that, compared with traditional symbol approach, our approach not only represents the key feature of time series, but also supports the similarity search effectively and has lower false positives rate.