Towards a faster symbolic aggregate approximation method

TL;DR

This paper introduces an enhanced SAX-based method for time series similarity search that incorporates an additional exclusion condition, significantly improving search speed over traditional SAX.

Contribution

The paper proposes a novel SAX-based approach with an added exclusion condition and dual representations to accelerate time series similarity searches.

Findings

The new method outperforms traditional SAX in speed.

Pre-computed distances enable efficient exclusion of non-matching series.

Experiments confirm the method's improved performance.

Abstract

The similarity search problem is one of the main problems in time series data mining. Traditionally, this problem was tackled by sequentially comparing the given query against all the time series in the database, and returning all the time series that are within a predetermined threshold of that query. But the large size and the high dimensionality of time series databases that are in use nowadays make that scenario inefficient. There are many representation techniques that aim at reducing the dimensionality of time series so that the search can be handled faster at a lower-dimensional space level. The symbolic aggregate approximation (SAX) is one of the most competitive methods in the literature. In this paper we present a new method that improves the performance of SAX by adding to it another exclusion condition that increases the exclusion power. This method is based on using two representations of the time series: one of SAX and the other is based on an optimal approximation of the time series. Pre-computed distances are calculated and stored offline to be used online to exclude a wide range of the search space using two exclusion conditions. We conduct experiments which show that the new method is faster than SAX.