next up previous contents
Next: Stationarity Up: PRACTICAL ASPECTS Previous: PRACTICAL ASPECTS   Contents


The dimension and entropy estimation methods require an infinite amount of noise-free data. In practice, the time series has finite length and precision. Furthermore, since an attractor has a finite size, it can only be studied at a certain distance range, the scaling region. The problem addressed in this chapter is to find optimal strategies for the estimation of the dimension and entropy of an attractor using an experimental time series.