Iskalni niz:
išči po
išči po
išči po
išči po
Vrsta gradiva:
Jezik:
Št. zadetkov: 15
Comparing solvability patterns of algorithms across diverse problem landscapes
Ana Nikolikj, Tome Eftimov
Objavljeni znanstveni prispevek na konferenci
Oznake: single-objective optimization;latent representations;explainability;
Leto: 2024 Vir: Institut Jožef Stefan (IJS)
Exploring module interactions in modular CMA-ES across problem classes
Ana Nikolikj, Tome Eftimov
Izvirni znanstveni članek
Oznake: module importance;empirical study;black-box optimization;benchmarking;
This study presents an in-depth analysis of module importance within the modular CMA-ES (modCMA-ES) algorithm using exploratory data analysis and large-scale benchmarking across the BBOB suite. Rather than introducing new algorithms, our contribution lies in uncovering how individual modules and the ...
Leto: 2025 Vir: Institut Jožef Stefan (IJS)
RF+clust for leave-one-problem-out performance prediction
Ana Nikolikj, Carola Doerr, Tome Eftimov
Objavljeni znanstveni prispevek na konferenci
Oznake: algorithm performance prediction;automated machine learning;zero-shot learning;black-box optimization;
Leto: 2023 Vir: Institut Jožef Stefan (IJS)
Recent advances in meta-features for automated single-objective black-box optimization
Gjorgjina Cenikj, Ana Nikolikj, Tome Eftimov
Video in druga učna gradiva
Oznake: black-box optimization;
Leto: 2024 Vir: Institut Jožef Stefan (IJS)
Customized exploration of landscape features driving multi-objective combinatorial optimization performance
Ana Nikolikj, Gabriela Ochoa, Tome Eftimov
Objavljeni znanstveni prispevek na konferenci
Oznake: landscape analysis;multi-objective combinatorial optimization;feature importance;
Customized exploration of landscape features driving multi-objective combinatorial optimization performance
Leto: 2025 Vir: Institut Jožef Stefan (IJS)
Recent advances in meta-features used for representing black-box single-objective continuous optimization
Gjorgjina Cenikj, Ana Nikolikj, Tome Eftimov
Objavljeni povzetek znanstvenega prispevka na konferenci
Oznake: learning algorithm features;learning problem landscape features;machine learning;
Recent advances in meta-features used for representing black-box single-objective continuous optimization
Leto: 2025 Vir: Institut Jožef Stefan (IJS)
Quantifying individual and joint module impact in modular optimization frameworks
Ana Nikolikj, Ana Kostovska, Diederick Vermetten, Carola Doerr, Tome Eftimov
Objavljeni znanstveni prispevek na konferenci
Oznake: metapodatki;strojno učenje;meta-learning;single-objective optimization;module importance;
Leto: 2024 Vir: Institut Jožef Stefan (IJS)
Generalization ability of feature-based performance prediction models
Ana Nikolikj, Ana Kostovska, Gjorgjina Cenikj, Carola Doerr, Tome Eftimov
Objavljeni znanstveni prispevek na konferenci
Oznake: metapodatki;strojno učenje;meta-learning;single-objective optimization;module importance;
This study examines the generalization ability of algorithm performance prediction models across various bench-mark suites. Comparing the statistical similarity between the problem collections with the accuracy of performance prediction models that are based on exploratory landscape analysis feature ...
Leto: 2024 Vir: Institut Jožef Stefan (IJS)
User-defined trade-offs in LLM benchmarking
Ana Gjorgjevikj, Ana Nikolikj, Barbara Koroušić-Seljak, Tome Eftimov
Izvirni znanstveni članek
Oznake: veliki jezikovni modeli;večkriterijsko odločanje;large language models;benchmarking;multi-criteria decision-making;
This paper presents xLLMBench, a transparent, decision-centric benchmarking framework that empowers decision-makers to rank large language models (LLMs) based on their preferences across diverse, potentially conflicting performance and non-performance criteria, e.g., domain accuracy, model size, ene ...
Leto: 2025 Vir: Institut Jožef Stefan (IJS)
Sensitivity analysis of RF+clust for leave-one-problem-out performance prediction
Ana Nikolikj, Michal Pluhacek, Carola Doerr, Peter Korošec, Tome Eftimov
Objavljeni znanstveni prispevek na konferenci
Oznake: strojno učenje;automated performance prediction;autoML;single-objective black-box optimization;zero-shot learning;
Leto: 2023 Vir: Institut Jožef Stefan (IJS)
Št. zadetkov: 15
Ključne besede:
Leto izdaje:
Avtorji:
Repozitorij:
Tipologija:
Jezik: