TIME SERIES ANALYSIS USING ENTROPY METRICS

Abstract

Entropy metrics are valuable tools for use in systems science and engineering. Both of domains are assumed to work with signals, to perform testing and evaluation tasks through specific signals. Entropy metrics indicate the degree of order/disorder in any given system, so they are capable to distinguish deterministic features from noise or chaos. The motivation of this paper is to present the main components and actual developments in entropy metrics used to analyze discrete time signals. The entropy metrics and measurements developed in this work try to find regularity in discrete time series in order to inspect them or predict future events.

Keywords

• discrete time series
• entropy
• entropy metrics
• sample entropy