Tomislav Dolic

Development of event-detection machine learning algorithms in the field of electrical car charging infrastructure. Deployment of latter as web-services or applications, for the purpose of field/live testing. Data handling realized by python library-environments as well as via interfaces to database-systems.

Development of an ML time-series forecast-model for the energy production of PV-facilities, based on weather-data forecasts. Implementation of the model as a web-service for the purpose of field-testing was achieved. It serves as a better replacement for a previously unflexible and hard-wired model with a lot of input-parameters.

Development of surveillance solutions for executive forces. Programming of microcontoller as well as setting up of monitoring devices for deployment. Monitoring of events within the surveillance environment. Programming Environments: python (QtPy), TeraTerm, VBA, Debian

Focus: Numerical Analysis, Simulation, Algorithmics, Statistics, Theoretical physics Master-Thesis: Point source stimulation of dendrites of a pyramidal cell: a computational study

Bachelor-Thesis: "Quadratic eigenvalue problem, Numerical treatment of sparse Matrices" Short Abstract: Within the discretization-process of fluid meachanics for example, quadratic Eigenvalue Problems arise. The underlying matrices are sparse, the use of iterative techniques is advisable.

Energietechnik und Leistungselektronik