A Methodology for Effective Surrogate Learning in Complex Optimization

2026-02-09

AI Summary

This work addresses the trade-off between fidelity and computational efficiency in surrogate modeling for complex optimization problems, such as urban traffic signal control. To this end, the authors propose the PTME methodology, which introduces a novel four-dimensional evaluation framework encompassing accuracy, time, memory, and energy consumption to systematically guide the development of deep learning–based surrogate models. By integrating these surrogate models into a new metaheuristic algorithm, PTME significantly enhances both optimization efficiency and decision quality in real-world urban scenarios. Experimental results demonstrate that PTME not only achieves superior performance in traffic signal optimization but also offers generalizable design principles applicable to a broad range of complex optimization tasks.

Technology Category

• Search and Optimization: Metareasoning and Metaheuristics
• Planning, Routing, and Scheduling: Optimization of Spatio-temporal Systems
• Machine Learning: Optimization

Application Category

• User Modeling, Personalization and Recommendation: User privacy protection in personalized systems
• Graph Algorithms and Modeling for the Web: Graph neural networks and deep learning approaches for Web-related graphs
• Search and Retrieval-Augmented AI: Web evaluation methodologies and metrics

Abstract

Solving complex problems requires continuous effort in developing theory and practice to cope with larger, more difficult scenarios. Working with surrogates is normal for creating a proxy that realistically models the problem into the computer. Thus, the question of how to best define and characterize such a surrogate model is of the utmost importance. In this paper, we introduce the PTME methodology to study deep learning surrogates by analyzing their Precision, Time, Memory, and Energy consumption. We argue that only a combination of numerical and physical performance can lead to a surrogate that is both a trusted scientific substitute for the real problem and an efficient experimental artifact for scalable studies. Here, we propose different surrogates for a real problem in optimally organizing the network of traffic lights in European cities and perform a PTME study on the surrogates'sampling methods, dataset sizes, and resource consumption. We further use the built surrogates in new optimization metaheuristics for decision-making in real cities. We offer better techniques and conclude that the PTME methodology can be used as a guideline for other applications and solvers.

Problem

• Surrogate Learning
• Complex Optimization
• Proxy Modeling
• Performance Evaluation
• Computational Efficiency

Innovation

• PTME methodology
• surrogate learning
• deep learning surrogates
• complex optimization
• resource-efficient modeling