Associate Professor · Department of Computer Science · Khmelnytskyi National University, Ukraine 🇺🇦
I build explanation methods that domain experts can check. The core idea is the transition matrix: an explicit mapping from the features a deep model has learned to the features a clinician or engineer already reasons with, so an explanation becomes a change of basis rather than a heatmap. I extend it with Lie-group equivariance and rough-set rules, and apply it in medical imaging and signals, UAV inspection of energy infrastructure, and smart mobility.
| Repository | What it implements | |
|---|---|---|
| 🫀 Medical XAI | transition-matrix-cardia-x | CARDIA-X – auditable post hoc ECG explainability (AI, 2026) |
| transition-matrix-cinetra | CINETRA – a registered, once-only audit of semantic concept transfer in cine cardiac MRI | |
| transition-matrix-dss | SEMTRA – global semantic transition and rough-set rules (MAKE, 2026) | |
| 🧠 Core XAI | transition-matrix-etm-xai | Equivariant transition matrices via Lie-group linearisation (MAKE, 2026) |
| transition-matrix-dl | Visual analytics with transition matrices (Mathematics, 2024) | |
| 🛩️ UAV & energy | my-study-slides | YOLO ensemble for UAV-based multispectral defect detection in wind turbines (IEEE IDAACS 2025, slides) |
Open course materials on GitHub:
- Intelligent Data Analysis – Jupyter curriculum with lectures, laboratory classes and course projects; bachelor's, in Ukrainian
- Conducting Experiments and Processing the Results of Scientific Research – PyTorch notebooks, assignments and lecture notes; master's, in English
- Statistical Data Processing – lecture notes, assignments and datasets; bachelor's, in Ukrainian
- Operations Research and Fundamentals of Decision Theory – tutorials; bachelor's, in Ukrainian
- Python Programming – an introductory course in step-by-step lessons, from variables and data structures to functions and modules
- CARDIA-X: Global Semantic Transition and Rough-Set Rules for Auditable Post Hoc Electrocardiographic Explainability. AI 7(9):376, 2026. doi:10.3390/ai7090376
- A Geo-consistent Weather Augmentation Method for Improving UAV Localization Robustness under Adverse Meteorological Conditions. Journal of Intelligent & Robotic Systems, 2026. doi:10.1007/s10846-026-02454-1
- Automated Synthesis of Hierarchical Deep Learning Cascades for Identifying Visually Similar Objects in UAV Imagery. Technologies 14(6):360, 2026. doi:10.3390/technologies14060360
Full list with citation badges: radiukpavlo.github.io/publications.
I welcome Horizon Europe, MSCA and EuroHPC consortia looking for an XAI task or work-package lead, a medical AI partner, UAV inspection of energy assets, or a Ukrainian pilot site in Khmelnytskyi – I am a researcher in the Horizon Europe project U_CAN. The partner kit has a capability statement, bios and our institution's PIC, and there is a one-page CV for proposal annexes.
Service: 450+ peer reviews (Web of Science) · programme committees of DEXA, WISE and MobiWis · organising committee of the ExplAI workshop.