Robertas Jurkus
Department: Image and Signal Analysis Group
Position: PhD student
Field: Informatics Engineering
Address: Akademijos st. 4, Vilnius
E-mail:
Doctoral studies
Theme
Maritime Traffic Awareness Evaluation Using Deep Neural Networks
Supervisor Assoc. Prof. Dr. Povilas Treigys
Study time 2020 10 01–2024 09 30
Aim
Investigate deep recurrent neural networks for monitoring the awareness of sea traffic - predicting the movement of a vessel.
Main problems:
- Application of recurrent neural network architectures for cargo vessel trajectory prediction.
- Study of hyperparameters of recurrent neural networks - determination of cell size.
- Application of different coordinate systems in identifying geolocation and their impact on forecasting.
- Introduction of different vessel types to the data set using the autoencoder architecture and quartile prediction evaluation.
Qualification
- 2020 - now, Vilnius University. Dissertation topic: Video, audio and streaming data processing technologies based on artificial intelligence;
- 2018 - 2020, Klaipėda University. Program: Engineering of technical information systems. Degree: master's degree in computer science;
- 2017 - 2018, Klaipėda University. Program: Informatics additional studies (to pursue a master's degree);
- 2013 - 2016, Klaipėda State College. Program: Informatics. Degree: Professional Bachelor's Degree in Computer Science.
- 2016 - now, work in UAB company, in the position of programmer.
Pedagogical activities
from 2022, Klaipėda University, lecturer:
- Java technologies
- Supervision of final bachelor's theses
2022, Klaipėda University, presenter of the ReactJS training program for front-end programmers
since 2020, Klaipėda University, assistant:
- Programming in .NET
- Development of embedded systems (Java)
- Reviewing theses
Scientific publications
- Jurkus R., Treigys P., Venskus J. (2021) Investigation of Recurrent Neural Network Architectures for Prediction of Vessel Trajectory. In: Lopata A., Gudonienė D., Butkienė R. (eds) Information and Software Technologies. ICIST 2021. Communications in Computer and Information Science, vol 1486. Springer, Cham. https://doi.org/10.1007/978-3-030-88304-1_16
Presentations, courses and seminars
In 2022:
- An article was prepared at the student scientific conference "Technology and Business Current Affairs" and a report was presented "The influence of vessel types on predicting the trajectory of vesel movement using deep recurrent neural networks". KTU, Panevėžys Faculty of Technology and Business.
- The presentation was read at the event "AI Lithuania Meetup Klaipėda", the title of the presentation was "Application of polar and Cartesian coordinate systems for trajectory prediction using recurrent neural networks", University of Klaipėda.
- Participated in the competition and read the report "The transport theme of students' final theses in the competition". The winner of the prize place, on the topic of water transport, entitled "Study of LSTM deep neural networks for predicting ship progress using big traffic data", Ministry of Communications, Vilnius, Lithuania.
- Participated in the event "Baltic Sea Science Day Award for Young Scientists 2022", research area: "Digital growth", topic: "Maritime Traffic Awareness Evaluation Using Deep Neural Networks", University of Latvia, Norwegian Presidency 2021-2022.
In 2021:
- A poster presentation was prepared, on the topic: "Prediction of vessels trajectory using different coordinate systems". DAMSS: 12th conference on data analysis methods for software systems, Druskininkai, Lithuania.
- The article was presented at the conference "Investigation of recurrent neural network architectures for prediction of vessel trajectory". Information and software technologies: 27th international conference, ICIST 2021, Kaunas, Lithuania.
In 2016:
- Participated in the international scientific-practical student conference and read the report "Modeling of Information Systems and Computer Equipment and Household Appliance Stores Business".
- Participated in the student conference and read the report "Incident Data Analysis Cube".