Aleksandra Knapińska

Aleksandra Knapińska is a doctoral researcher at the Wrocław University of Science and Technology, Faculty of Electronics, Photonics and Microsystems, Department of Computer Networks and Systems. She is an active member of multiple research teams including the Computer Networks Team, Machine Learning Team, and Advanced Data Analysis Methods Team. Her work is primarily associated with the MAAN project focused on optimizing multilayer application-aware networks.

Her research interests center on network optimization, machine learning for traffic prediction, and multi-criteria optimization in optical and multilayer networks. She applies advanced data analysis and machine learning techniques to model and predict time-varying network traffic, particularly in backbone and optical networks. Her work bridges theoretical modeling with practical network performance evaluation.

The most recent articles highlight a strong trend in using machine learning—especially ensemble and neural network models—for short-term and long-term network traffic forecasting. These publications emphasize challenges such as feature selection, bandwidth blocking, node failure resilience, and traffic fragmentation in spectrally-spatially flexible optical networks. The research is deeply applied, targeting real-world network performance improvements.

Scientific profiles:

• Google Scholar
• ResearchGate
• ORCID: 0000-0003-2654-4893

She is involved in supervising diploma theses and has participated in international research collaborations, including an internship at Politecnico di Torino, Italy, under the InterDocSchool program. She has presented her work at major conferences such as GLOBECOM, ONDM, RNDM, and CNSM. Her PhD thesis, titled "Optimization of multilayer networks with time-varying traffic aided by traffic prediction", is supervised by Prof. Krzysztof Walkowiak, with Dr. Piotr Lechowicz as co-supervisor.

She is part of ongoing research projects including:

• Dark-Box Optimization – Developing highly effective general-purpose optimizers
• Evolutionary methods for multi-criteria optimization with many criteria
• Advanced optimization of multi-layer application-aware networks (MAAN)
• Using multi-criteria optimization in classifier training for decision-making tasks