Visiting Scientist from ASPETE, School of Pedagogical and Technological Education
nikmatz@gmail.com
Friedrich-Alexander-Universität Erlangen Nürnberg. FAU DCN-AvH Chair for Dynamics, Control, Machine Learning and Numerics – Alexander von Humboldt Professorship.
+49 9131 85-67170
I am an Associate Professor of Applied Mathematics and Artificial Intelligence in Education at the School of Pedagogical and Technological Education in Athens, Greece.
My early research focused on nonlinear analysis and differential inclusions, particularly on boundary value problems involving multivalued nonlinearities, maximal monotone operators, and hemivariational structures. In my doctoral work at the National Technical University of Athens, I studied nonlinear boundary value problems for ordinary, elliptic, and evolution differential inclusions, developing existence results using methods from nonlinear functional analysis, variational techniques, and multivalued analysis.
In later years, my research interests expanded toward mathematics education and technology-enhanced learning, with particular emphasis on the use of computational tools and digital environments in the teaching of mathematics. More recently, I have focused on the integration of generative artificial intelligence (GenAI) tools in mathematics and engineering education, exploring how large language models and AI-based assistants can support mathematical reasoning, problem solving, and inquiry-based learning. This work also examines how computer algebra systems, generative AI, and interactive digital platforms can be combined to design modern mathematics laboratory environments and technology-supported learning experiences in higher education.
More recently, my research has explored the interaction between applied mathematics, data-driven modeling, and machine learning. In particular, I investigate the use of neural networks and neural ordinary differential equations (Neural ODEs) as emerging frameworks for modeling complex dynamical systems. My work focuses on the mathematical properties and dynamical behavior of these models, including issues related to stability, sensitivity, and the role of model architecture and activation functions in the representation of nonlinear dynamics. This line of research aims to combine tools from dynamical systems and nonlinear analysis with modern machine learning approaches for the study and modeling of complex time-dependent phenomena.
