Similarity-based Portfolio Construction for Black-box Optimization

2026-04-20

AI Summary

This work addresses the instability of single-algorithm approaches in black-box optimization, which often suffer from selection bias or high variance. To overcome this limitation, the authors propose a k-nearest neighbors-based dynamic algorithm portfolio method that constructs a customized set of algorithms for each new problem instance by leveraging similarity matching with previously observed instances. Within a fixed evaluation budget, the method adaptively allocates computational resources to harness the complementary strengths of multiple algorithms. Moving beyond the conventional paradigm of selecting a single solver, the proposed approach not only significantly outperforms the strongest baseline—the virtual best solver—but also achieves markedly improved solution quality and robustness in a single run.

Technology Category

• Search and Optimization: Sampling/Simulation-based Search
• Constraint Satisfaction and Optimization: Solvers and Tools
• Reasoning under Uncertainty: Stochastic Optimization

Application Category

• Graph Algorithms and Modeling for the Web: Efficient manipulation of static and dynamic Web-related graphs
• Responsible Web: Human-perceived consequences of algorithmic deployment on the web
• User Modeling, Personalization and Recommendation: Attacks and countermeasures in recommendation systems

Abstract

In black-box optimization, a central question is which algorithm to use to solve a given, previously unseen, problem. Selecting a single algorithm, however, entails inherent risks: inaccuracies in the selector may lead to poor choices, and even well-performing algorithms with high variance can yield unsatisfactory results in a single run. A natural remedy is to split the evaluation budget across multiple runs of potentially different algorithms. Such sequential algorithm portfolios benefit from variance reduction and complementarities between algorithms, often outperforming approaches that allocate the entire budget to a single solver. While effective portfolios can be constructed post-hoc, transferring this idea to the algorithm selection setting is non-trivial. We show that a naive portfolio constructed over the full training set already outperforms the strongest traditional baseline, the virtual best solver. We then propose a simple yet effective k-nearest-neighbor-based finetuning approach to construct portfolios tailored to unseen instances, yielding further improvements and highlighting the effectiveness of portfolio selection in fixed-budget black-box optimization.

Problem

• black-box optimization
• algorithm selection
• portfolio construction
• fixed-budget optimization
• variance reduction

Innovation

• algorithm portfolio
• black-box optimization
• k-nearest neighbors
• instance similarity
• fixed-budget optimization