Algorithm Selection | Intelligent Optimization Lab

Large Language Models for Optimization

Sat, 26 Sep 2026 00:00:00 +0000

IOLab investigates how large language models can complement established combinatorial optimization techniques. Rather than treating an LLM as a stand-alone solver, we study its role as one component of a structured optimization system.

Our work examines LLM-based generative operators within metaheuristics, the ability of models to recall and interpret problem features, and the information encoded in their internal representations. These representations can be evaluated as surrogates for predicting instance difficulty and algorithm behavior in algorithm-selection scenarios.

The group has published a systematic review of LLMs for combinatorial optimization and experimental work on feature extraction and algorithm selection.

Behavior and representation in open-weight Large Language Models for combinatorial optimization: From feature extraction to algorithm selection

Thu, 01 Jan 2026 00:00:00 +0000

Add the full text or supplementary notes for the publication here using Markdown formatting.

Automated Algorithm Selection in Optimization and Machine Learning

Thu, 05 Oct 2023 09:30:00 +0200

Nysret Musliu is with the Christian Doppler Laboratory for AI and Optimization in Planning and Scheduling at the Institute of Logic and Computation, Technische Universität Wien, Austria.

The course is organized in three sessions:

5 October 2023, 9:30-12:30
5 October 2023, 14:00-17:00
6 October 2023, 9:30-11:30

Industrial Production Scheduling

Sun, 01 Jan 2017 00:00:00 +0000

IOLab develops optimization methods for complex production environments, where scheduling decisions must account for technological constraints, limited resources, energy consumption, and operational objectives.

From 2017 to 2021, the group conducted a research collaboration with Danieli Automation on production scheduling for steel plants. Led by Luca Di Gaspero, the project studied metaheuristic methods for industrial decision-support systems in an Industry 4.0 setting.

More recent work investigates the Oven Scheduling Problem and related parallel-batch scheduling models. The group has developed local-search and large-neighborhood-search methods, as well as instance-space analysis and algorithm-selection techniques for understanding when different algorithms perform best.

Code and experimental data are available through the OSP-LS and OSP-LNS repositories.