Research interest
My research focuses on the design, implementation, and rigorous experimental analysis of algorithms for combinatorial optimization. I am particularly interested in metaheuristic methods, their parallelization on modern computing architectures, and the development of systematic evaluation methodologies that provide deeper insights into algorithmic behavior beyond sdandard metrics.
Combinatorial Optimization and Scheduling
- Modeling and solving hard (NP-hard) optimization problems such as scheduling and related resource-allocation settings.
- Identifying problem features that influence difficulty and algorithm performance.
- Designing benchmark-driven computational studies to distinguish easy vs. hard instances and to study performance variability.
Metaheuristics and Algorithm Design
- Design and refinement of metaheuristics, with a strong focus on Variable Neighborhood Search (VNS) and Bee Colony Optimization method (BCO).
- Development and analysis of neighborhood structures, diversification/intensification mechanisms, and hybrid strategies.
- Comparative evaluation of heuristic strategies, including careful ablation studies and principled reporting of results.
Parallel Computing and High-Performance Optimization
- Shared-memory and distributed-memory parallelization strategies for metaheuristics (e.g., parallel neighborhood evaluation, multi-search, cooperation schemes).
- Implementation-oriented research: scalability, synchronization/communication trade-offs, and performance profiling.
- Reproducible computational experimentation pipelines for large-scale runs and systematic comparisons.
Systematic Literature Reviews and Taxonomies
- Systematic organization of research fields through classifications and taxonomies of algorithmic and implementation strategies.
- Evidence-driven synthesis of results across studies, highlighting recurring patterns, gaps, and methodological pitfalls.
- Bridging conceptual contributions and empirical evidence to support clear, defensible conclusions.
Data Analysis and Visualization for Algorithmic Insight
- Statistical analysis of computational results to quantify the influence of instance parameters and algorithmic design choices.
- Visualization of performance distributions, sensitivity, and interactions to support transparent interpretation.
- Tooling: R for data analysis and visualization; C/C++ for scientific computing and high-performance implementations.