Times series discretization using evolutionary programming

Abstract

In this work, we present a novel algorithm for time series discretization. Our approach includes the optimization of the word size and the alphabet as one parameter. Using evolutionary programming, the search for a good discretization scheme is guided by a cost function which considers three criteria: the entropy regarding the classification, the complexity measured as the number of different strings needed to represent the complete data set, and the compression rate assessed as the length of the discrete representation. Our proposal is compared with some of the most representative algorithms found in the specialized literature, tested in a well-known benchmark of time series data sets. The statistical analysis of the classification accuracy shows that the overall performance of our algorithm is highly competitive. © 2011 Springer-Verlag.