An entropy type measure of complexity

摘要

In a continuous system the Shannon entropy is defined as the expected value of the information content of a random variable X. In this paper we assume that the random variable X comes from the generalized γ-ordered Normal distribution N_γ which can be used in applications concerning heavy-tailed distributions. The generalized Shannon entropy is also introduced through the generalized Fisher's information. In this context, a useful measure for technological applications is introduced and studied, which extends the SDL measure of complexity used in the study of the EEG signals on epileptic seizures.