An improved estimator of Shannon entropy with applications to systems with memory

Abstract

We investigate the memory properties of discrete sequences built upon a finite number of states. We find that the block entropy can reliably determine the memory for systems modeled as Markov chains of arbitrary finite order. Further, we provide an entropy estimator that remarkably gives accurate results when correlations are present. To illustrate our findings, we calculate the memory of daily precipitation series at different locations. Our results are in agreement with existing methods being at the same time valid in the undersampled regime and independent of model selection.

References in corpus (5)

• Bias Analysis in Entropy Estimation
• Detecting Memory and Structure in Human Navigation Patterns Using Markov Chain Models of Varying Order
• Optimal non-Markovian search strategies with n-step memory
• Regular and stochastic behavior of Parkinsonian pathological tremor signals
• Additive N-Step Markov Chains as Prototype Model of Symbolic Stochastic Dynamical Systems with Long-Range Correlations