Event details from: Permutation Entropy: Extension to Graph Signals and Application Scenarios

Date
19 November 2026
Time
11:00 – 12:00
Room
The second-floor library

In recent years, entropy measures have been widely employed in a variety of application scenarios. In particular, permutation entropy (PE), which is based on the relative frequencies of ordinal patterns identified along a sequence of samples, enables the characterization of time series and, consequently, the classification of signals in different contexts. In this talk, I discuss the extension of PE to signals defined on graphs, which are commonly used to model data arranged over networks. The computation of graph-based permutation entropy aims to capture the interdependencies among signal samples, as determined by the topology of the underlying graph. The potential benefits of this approach are illustrated through preliminary applications involving signals collected from both artificial and biological systems.

About the speaker

Juliano Bandeira Lima is an Associate Professor in the Department of Electronics and Systems at the Federal University of Pernambuco (UFPE-Brazil), where he also teaches in the Graduate Programme in Electrical Engineering and leads the Digital Signal Processing Research Group. He received his BSc, MSc, and PhD in Electrical Engineering from UFPE, including a doctoral research period at Carleton University in Canada, and completed a postdoctoral fellowship at the University of Minho’s School of Medicine in Portugal in 2025–2026. His research focuses on signal-processing methods and their applications to communications, neuroscience, and cryptography. He is a Senior Member of the IEEE and a CNPq Research Productivity Fellow.