Authors: Xiao-Hua Lv, Yong-Sheng Xue, Zi-Yu Lin
Tags: 2004, conceptual modeling
We target at the growing topic of representing and searching time-series data. A new MABI (Moving Average Based Indexing) technique is proposed to improve the performance of the similarity searching in large time-series databases. Notions of Moving average and Euclidean distances are introduced to represent the time-series data and to measure the distance between two series. Based on the distance reducing rate relation theorem, the MABI technique has the ability to prune the unqualified sequences out quickly in similarity searches and to restrict the search to a much smaller range, compare to the data in question. Finally the paper reports some results of the experiment on a stock price data set, and shows the good performance of MABI method.Read the full paper here: https://link.springer.com/chapter/10.1007/978-3-540-30466-1_27
