Abstract:The performance of automated algorithm selection (AAS) strongly depends on the portfolio of algorithms to choose from. Selecting the portfolio is a non-trivial task that requires balancing the trade-off between the higher flexibility of large portfolios with the increased complexity of the AAS task. In practice, probably the most common way to choose the algorithms for the portfolio is a greedy selection of the algorithms that perform well in some reference tasks of interest. We set out in this work to investigate alternative, data-driven portfolio selection techniques. Our proposed method creates algorithm behavior meta-representations, constructs a graph from a set of algorithms based on their meta-representation similarity, and applies a graph algorithm to select a final portfolio of diverse, representative, and non-redundant algorithms. We evaluate two distinct meta-representation techniques (SHAP and performance2vec) for selecting complementary portfolios from a total of 324 different variants of CMA-ES for the task of optimizing the BBOB single-objective problems in dimensionalities 5 and 30 with different cut-off budgets. We test two types of portfolios: one related to overall algorithm behavior and the `personalized' one (related to algorithm behavior per each problem separately). We observe that the approach built on the performance2vec-based representations favors small portfolios with negligible error in the AAS task relative to the virtual best solver from the selected portfolio, whereas the portfolios built from the SHAP-based representations gain from higher flexibility at the cost of decreased performance of the AAS. Across most considered scenarios, personalized portfolios yield comparable or slightly better performance than the classical greedy approach. They outperform the full portfolio in all scenarios.

A Learning-Based Algorithm Selection Meta-Reasoner for the Real-Time MPE Problem

Deep Reinforcement Learning for Dynamic Algorithm Selection: A Proof-of-Principle Study on Differential Evolution

Automated Selection of Evolutionary Multi-objective Optimization Algorithms.

Time series model selection with a meta-learning approach; evidence from a pool of forecasting algorithms

Learning to select computations

A many-objective integrated evolutionary algorithm for feature selection in anomaly detection

Effective Response Strategies Based on Adaptive Selection for Dynamic Multi-Objective Evolutionary Optimization

Anytime Automatic Algorithm Selection for the Multi-Agent Path Finding Problem

EFFECT: Explainable Framework for Meta-Learning in Automatic Classification Algorithm Selection

PS-AAS: Portfolio Selection for Automated Algorithm Selection in Black-Box Optimization

Multi-Objective Model Selection Via Racing

MetaSelector: Meta-Learning for Recommendation with User-Level Adaptive Model Selection

Per-Instance Algorithm Selection for Recommender Systems via Instance Clustering

Meta-Learning from Learning Curves for Budget-Limited Algorithm Selection

Improving Algorithm-Selection and Performance-Prediction via Learning Discriminating Training Samples

Genetic Optimization of Meta-Learning Schemes for Context-Based Fault Detection

One-at-a-time: A Meta-Learning Recommender-System for Recommendation-Algorithm Selection on Micro Level

A new learning-based adaptive multi-objective evolutionary algorithm.

Learning Heuristic Selection with Dynamic Algorithm Configuration

A meta-learning framework for algorithm recommendation in software fault prediction

A data-driven meta-learning recommendation model for multi-mode resource constrained project scheduling problem