A new method of piecewise linear approximation of non-stationary time series for similarity measurement

As much of data and information are performed over time they are represented in a form of time series. The amount of information increases with tremendous speed that should be stored and analyzed. Extracting meaningful information from the stock of data is called time series data mining. Important part of time series data mining is similarity search of time series. The new method is developed for time series representation in a form of piecewise linear form which was taken as a base for building as criteria for similarity search of time series. The main approach based on our work will be discussed in this article.