Error Restricted Piecewise Linear Representation of Time Series Based on Special Points
Pengtao Jia, Huacan He, Tao Sun · 2008
Time series composed of a series of data observed according to time sequence. There are large numbers of time series in the field of automatic control. Usually time series data is massive. Sometimes, it is inefficient or impractical to mine the original time series data directly. So time series should be represented on higher level. Piecewise Linear Representation is perhaps the most frequently used representation with lower dimensions of index structure and the faster speed of calculation. But PLR needs a parameter k given by users and to make certain k is not easy for the users. We put forward an algorithm of error restricted piecewise linear representation based on special points that can give the number of fitting lines automatically according to a given error e. And it has lower time complexity than other error restricted algorithm, such as sliding windows, Bottom-Up, Top-Down. According to experiments our algorithm leads to less fitting error and better performance than other algorithms. So it would be more efficient in the real application of control fields.