Research on Similarity Search Technique of Variable Long Time Series Data Mining

Abstract

Time series similarity search is the main subroutine of time series data mining algorithm. The efficiency of time series similarity research has become an obstacle to the development of time series mining algorithms. The representation of time series and the measurement of similarity are the basis of time series similarity research and play a crucial role in completing the similarity search task of time series. As a method of measuring similarity, the dynamic distortion of time can be effectively dealt with by deforming the time series over time, and it has good stability. However, time series data is usually an ever-increasing data stream, and the direct study of similarity will cause considerable storage space consumption, and may affect the accuracy and reliability of the algorithm. Therefore, it is necessary to determine the time series in advance, express the main self of the original time series in a concise and abstract form, and carry out similarity search on the developed sequence to improve the similarity search efficiency of the sequence.