School of Electrical and Computer Engineering
2310995955
My office is located at the central building of the Faculty of Engineering facing Egnatia street (gray pillars marble steps), just go straight to the end of the hall (do not turn right to the corridor), take the stairs, at the first floor, next to the three elevators. The new collaborators and members of the research team are hosted also there.
Dimitris Kugiumtzis is Professor on Computational Statistics – Time Series at the Department of Electrical and Computer Engineering (ECE). He was Associate Professor at ECE (2013-2017), Assistant and Associate Professor at the Department of Mathematical, Physical and Computational Sciences (2001-2013), Lecture B at the Department of Statistics, University of Glasgow (2000-2001), guest scientist (PostDoc) at the Max-Planck-Institute for Physics of Complex Systems in Dresden (1998-1999) and associate researcher at the State Center for Epilepsy in Norway (1997). He is director of the Informatics Lab ECE since 2019 and the Interdepartmental MSc on Biomedical Engineering (2021-2023). He supervised 8 PhD theses (other two in progress), over 40 MSc theses and 20 diploma theses. He has 95 peer-reviewed publications, 24 publications in international and 29 in national conference proceedings. He has h-index 27 and 2226 citations (source Web of Science, 11/5/2023).
11/5/2023 Maximal Spectral Overlap Wavelet Transform (MSO-WT). The software in Python for the proposed method MSO-WT for adaptive decomposition of multicomponent Signals and estimation of phase synchronization, generalizing the standard Mean Phase Coherence (MPC) to Multi-Component Mean Phase Coherence (MCMPC), can be found in github, https://github.com/apostolosev/MSO_WT. The measure is presented in the paper: A. Evangelidis, D. Kugiumtzis, “Adaptive Decomposition of Multicomponent Signals and Estimation of Phase Synchronization“, IEEE Transactions on Signal Processing, early access, 2023.
17/1/2021 Diverse standard measures of Granger causality: Granger causality index (GCI), conditional Granger causality index (CGCI), transfer entropy (TE), Partial Transfer Entropy (PTE). The measures are developed in Matlab and given in github, github, https://github.com/dkugiu/Matlab/tree/master/GrangerCausalityMeasures . Please cite the paper when appropriate: E. Siggiridou, Ch. Koutlis, A. Tsimpiris, D. Kugiumtzis, “Evaluation of Granger Causality Measures for Constructing Networks from Multivariate Time Series“, Entropy, Vol 21 (11): 1080, 2019.
17/1/2021 PTERV: The Matlab program for the direct causality measure of partial transfer entropy on ranked vectors is in github, https://github.com/dkugiu/Matlab/blob/master/PTERV.m . The measure is presented in the paper: D. Kugiumtzis, “Partial Transfer Entropy on Rank Vectors“, The European Physical Journal Special Topics, Vol 222, No 2, pp 401-420, 2013
25/3/2019, Plug-in algorithm for filling the gap of transcranial magnetic stimulation (TMS) artifact in the electroencephalogam (EEG). Author: Alexandra Anagnostopoulou (developed in the frame of the diploma thesis in the Dep of Electrical and Computer Engineering, AUTh). Two Matlab plug-ins are given in two respective zip files: for the Matlab module TESA, FillTMSgapTESA.zip; for the Matlab module TMSEEG (website: http://www.tmseeg.com/ ), FillTMSgapTMSEEG.zip.
4/8/2015, RCGCI and mBTS: The software in Matlab developed together with Elsa Siggiridou for the restricted conditional Granger causality index (RCGCI) based on the dynamic regression model derived by the modified backward-in-time selection (mBTS) algorithm. The folder in github, https://github.com/dkugiu/Matlab/tree/master/RCGCI, contains an example script and the functions implementing the algorithm (see first the file README.txt). The measure is presented in the paper: E. Siggiridou, D. Kugiumtzis, “Granger Causality in Multi-variate Time Series using a Time Ordered Restricted Vector Autoregressive Model“, IEEE Transactions on Signal Processing, Vol 64(7), pp 1759-1773, 2016.
Updated 3/4/215, PMIME: The software in Matlab for the direct causality measure of Partial Mutual Information from Mixed Embedding (PMIME) is given in github, https://github.com/dkugiu/Matlab/tree/master/PMIME. The measure is presented in the paper: D. Kugiumtzis, “Direct coupling information measure from non-uniform embedding“, Physical Review E, Vol 87, 062918, 2013.
Last updated MATS: Together with Alkiviadis Tsimpiris we have developed the Measures of Analysis of Time Series (MATS) open source matlab toolkit. You may download MATS from http://eeganalysis.web.auth.gr.
You can also download it from the web-page of the Journal of Statistical Software (free access) at http://www.jstatsoft.org/v33/i05/ .You can also find the paper presenting MATS there.
Last updated, STAP: The software in Matlab for the generation of surrogate time series with the algorithm of Statistically Transformed Autoregressive Process (STAP) is here. The measure is presented in the paper: D. Kugiumtzis, “Statically Transformed Autoregressive Process and Surrogate Data Test for Nonlinearity“, Physical Review E, Vol 66, 025201, 2002.
Probability Theory and Statistics (Undergraduate)
Data Analysis (Undergraduate)
Time Series (Undergraduate)
Data Analysis and Processing in Matlab (MSc in Computational Physics, Physics Dept) (Postgraduate)
Time Series Analysis (MSc in Statistics and Modeling, Mathematics Dept) (Postgraduate)
Biomedical Data Acquisition and Signal Processing (MSc Biomedical Engineering, 1/4 of the course) (Postgraduate)
Computational neuroscience – neuroengineering (MSc Biomedical Engineering, 1/3 of the course) (Postgraduate)
Introduction to Computational Neuroscience (MSc in Advanced Systems of Communication and Computers, 1/2 of the course) (Postgraduate)
Big Data Analytics (MSc in Advanced Systems of Communication and Computers, 1/2 of the course) (Postgraduate)
Statistical Analysis of Networks (MSC in Networks and Complexity, 1/2 of the course) (Postgraduate)