Pitfalls and best practices in algorithm configuration
From MaRDI portal
Abstract: Good parameter settings are crucial to achieve high performance in many areas of artificial intelligence (AI), such as propositional satisfiability solving, AI planning, scheduling, and machine learning (in particular deep learning). Automated algorithm configuration methods have recently received much attention in the AI community since they replace tedious, irreproducible and error-prone manual parameter tuning and can lead to new state-of-the-art performance. However, practical applications of algorithm configuration are prone to several (often subtle) pitfalls in the experimental design that can render the procedure ineffective. We identify several common issues and propose best practices for avoiding them. As one possibility for automatically handling as many of these as possible, we also propose a tool called GenericWrapper4AC.
Recommendations
- Paramils: an automatic algorithm configuration framework
- Effect of transformations of numerical parameters in automatic algorithm configuration
- MADS/F-Race: mesh adaptive direct search meets F-Race
- Capping methods for the automatic configuration of optimization algorithms
- scientific article; zbMATH DE number 5829047
Cites work
Cited in
(15)- Are Humans Bayesian in the Optimization of Black-Box Functions?
- A causal framework for stochastic local search optimization algorithms
- A biased random‐key genetic algorithm for the home health care problem
- Effect of transformations of numerical parameters in automatic algorithm configuration
- Sequential model based optimization of partially defined functions under unknown constraints
- On the configuration of more and less expressive logic programs
- Benchmarking
- GenericWrapper4AC
- Modelling human active search in optimizing black-box functions
- Learning Enabled Constrained Black-Box Optimization
- MADS/F-Race: mesh adaptive direct search meets F-Race
- Automated Reinforcement Learning (AutoRL): A Survey and Open Problems
- Peeking beyond peaks: challenges and research potentials of continuous multimodal multi-objective optimization
- Paramils: an automatic algorithm configuration framework
- Efficient benchmarking of algorithm configurators via model-based surrogates
Describes a project that uses
Uses Software
This page was built for publication: Pitfalls and best practices in algorithm configuration
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q5376630)
