TEMPORAL PATTERN IDENTIFICATION OF TIME SERIES DATA USING PATTERN WAVELETS AND GENETIC ALGORITHMS

A new method for temporal pattern matching of a time series is developed using pattern wavelets and genetic algorithms. The pattern wavelet is applied to the matching of an embedded time series. A problem-specific fitness factor is introduced in the new algorithm, which is useful to construct a fitness function of the feature space. A two-step process discovers the pattern wavelet that yields high fitness value. The best temporal pattern matches are found through a thresholding process. These matches are kept and the future time series data point is used in the genetic algorithm's fitness function. The algorithm has been successfully applied to the identification of statistically significant temporal patterns in financial time series data.