This study introduces approximate entropy (ApEn) to assess the randomness of finite sequences, a gap in traditional probability theory. ApEn quantifies sequence irregularity, offering a novel approach to understanding randomness in real-world data.
Area of Science:
Information Theory
Probability Theory
Statistical Analysis
Background:
Axiomatic probability theory does not adequately address the assessment of randomness for finite sequences.
Assessing randomness in single, finite sequences is a ubiquitous but unaddressed problem.
Classical measures of randomness have limitations, especially for finite sequences.
Purpose of the Study:
To introduce and validate approximate entropy (ApEn) as a computable measure for assessing the randomness of single, finite sequences.
To demonstrate the utility of ApEn and associated deficit functions (def(m)) in refining concepts of probabilistic independence and normality.
To explore the application of ApEn to mathematical constants and sequences derived from irrational numbers.
Main Methods:
Application of approximate entropy (ApEn), a measure of sequential irregularity, to single sequences of finite and infinite length.
Identification of maximally irregular sequences for finite, finite-state sequences.
Development and application of deficit (def(m)) functions to quantify deviations from maximal irregularity.
Analysis of sequences including digits of mathematical constants (e, pi, radical2, radical3) and sequences from fractional parts of irrational multiples.
Main Results:
Approximate entropy (ApEn) provides a novel, multidimensional approach to assessing sequence randomness, applicable to finite sequences.
Deficit functions (def(m)) refine notions of probabilistic independence and normality, demonstrating utility with mathematical constants and irrational sequences.
Analytic results confirm the role and validity of axiomatic probability properties for specified sequences in the physical world.
Conclusions:
Approximate entropy (ApEn) offers a practical and effective method for evaluating the randomness of finite sequences, addressing a key limitation in probability theory.
The developed deficit functions provide a quantitative measure of deviation from maximal irregularity, enhancing the analysis of sequence properties.
The study bridges theoretical probability with practical sequence analysis, offering insights applicable to real-world data and mathematical sequences.