A Piecewise Aggregation Approximation of Time Series Based on Wavelet Entropy
Xin Wei Zheng · Jisuanji fangzhen · 2015
Due to the high dimensionality of time series,mining this kind of data directly is not very easy. The job we usually do before data mining is feature representation,in order to achieve the goal of dimensionality reduction. Piecewise aggregation approximation( PAA) is one of the most commonly used methods of feature representation. The PAA algorithm treats each section on average,this is one of the shortcomings of this algorithm. Aiming at this shortcoming,we presented a method based on wavelet entropy and piecewise aggregation approximation in this paper. We applied wavelet entropy algorithm to the improvement of the PAA,and treated wavelet energy entropy as an indicator of the complexity of one section. According to the wavelet entropy value of each section to allocate points within each section,we achieved a detailed description of the complexrange and a rough approximation of the relatively stable range. Finally,the Matlab simulation results show that this method in the same compression ratio can fit the original sequence better than other methods,it can not only effectively reduce the dimension of time series,and also make more accurate approximation,thus achieving the efficiency of time-series data mining.