Mr. Dimitrios Gerontitis | Neural Networks | Best Researcher Award

Dimitrios Gerontitis at International Hellenic University, Greece

Dimitrios Gerontitis is a Greek mathematician and researcher with a strong academic background in mathematics, theoretical informatics, and systems & control theory. With over a decade of experience in academia and scientific publishing, he has contributed significantly to international research through reviewing for high-impact journals and collaborating with global institutions.

Professional Profile:

Scopus

Orcid

Google Scholar

Education Background

• High School Diploma – Graduated with Excellence (18.8/20), 2008

• B.Sc. in Mathematics, Aristotle University of Thessaloniki (2008–2013)

  • Graduated with 7.32/10; Ranked 29th out of 224 graduates

• M.Sc. in Theoretical Informatics and Systems & Control Theory, Aristotle University of Thessaloniki (2013–2016)

  • Graduated with Excellence (9.3/10)

Professional Development

• Teaching Assistant in Mathematics I & II, Department of Information and Electronic Engineering, International Hellenic University (2023–2024)

• Reviewer for 25+ international scientific journals including IEEE, Elsevier, Springer, and Taylor & Francis (2018–Present)

• Math Educator, Professional Mathematics Tutor (2014–2015)

• Army Service: Operated digital terminals and crypto-machines (2016–2017)

• Research Collaborator, University of Bremen, DAAD-funded project on multipole techniques and EELS applications (2018)

Research Focus

• Recurrent Neural Networks

• Matrix Theory

• Numerical Linear Algebra

• Simulink Modeling

• Optimization Methods

Author Metrics:

Awards and Honors:

• Novel Discrete-Time Recurrent Neural Networks Handling Discrete-Form Time-Variant Multi-Augmented Sylvester Matrix Problems and Manipulator Application
  Y. Shi, L. Jin, S. Li, J. Li, J. Qiang, D.K. Gerontitis
  IEEE Transactions on Neural Networks and Learning Systems, Vol. 33, No. 2, pp. 587–599, 2020
  Citations: 66
  ➤ Developed novel discrete-time RNNs for solving advanced matrix problems with application to robotic manipulators.

• Gradient Neural Network with Nonlinear Activation for Computing Inner Inverses and the Drazin Inverse
  P.S. Stanimirović, M.D. Petković, D. Gerontitis
  Neural Processing Letters, Vol. 48, No. 1, pp. 109–133, 2018
  Citations: 45
  ➤ Proposed a gradient-based neural network architecture for generalized matrix inverse computations.

• Conditions for Existence, Representations, and Computation of Matrix Generalized Inverses
  P.S. Stanimirović, M. Ćirić, I. Stojanović, D. Gerontitis
  Complexity, Vol. 2017, Article ID 6429725
  Citations: 35
  ➤ Provided a comprehensive theoretical framework for the computation and representation of matrix generalized inverses.

• A Family of Varying-Parameter Finite-Time Zeroing Neural Networks for Solving Time-Varying Sylvester Equation and its Application
  D. Gerontitis, R. Behera, P. Tzekis, P. Stanimirović
  Journal of Computational and Applied Mathematics, Vol. 403, Article 113826, 2022
  Citations: 31
  ➤ Introduced a new family of finite-time neural network models for dynamic matrix equation solving.

• A Higher-Order Zeroing Neural Network for Pseudoinversion of an Arbitrary Time-Varying Matrix with Applications to Mobile Object Localization
  T.E. Simos, V.N. Katsikis, S.D. Mourtas, P.S. Stanimirović, D. Gerontitis
  Information Sciences, Vol. 600, pp. 226–238, 2022
  Citations: 28
  ➤ Proposed an innovative higher-order model for time-varying matrix pseudoinversion, aiding real-time object localization.

Conclusion