Engineering Optimization

Learning-Augmented Optimization: Theory, Algorithms, and Applications for Complex Decision-Making

Optimization has long been the quantitative backbone of engineering decision-making, with metaheuristics, mathematical programming, and robust and stochastic formulations providing the means to plan, design, and operate complex systems. In recent years, the rapid maturation of machine learning, predictive modelling, and reinforcement learning has begun to reshape how these optimization methods are constructed and deployed. Rather than replacing established optimization paradigms, learning techniques are increasingly used to strengthen them—guiding search, accelerating convergence, approximating expensive evaluations, and adapting solutions to uncertain and dynamic conditions. This emerging synthesis, in which learning augments rather than supplants optimization, opens a distinct and timely line of inquiry that the present Special Issue seeks to consolidate.

The central theme of this Special Issue is the principled integration of learning into optimization, where the optimization process remains the core object of study and learning serves to enhance its efficiency, scalability, robustness, and solution quality. Of particular interest are learning-guided metaheuristics, the coupling of prediction and optimization, the integration of reinforcement learning with mathematical programming and combinatorial optimization, surrogate-assisted optimization for computationally expensive problems, and data-driven approaches to robust, stochastic, and multi-objective optimization under uncertainty. Contributions that strengthen the theoretical foundations of such methods—including convergence, generalization, approximation guarantees, and post-optimality analysis—are equally encouraged.

The Special Issue welcomes original research spanning the full spectrum from theory to practice, including new algorithms and analytical results, rigorous empirical studies on challenging benchmark and real-world instances, and substantive engineering applications that demonstrate practical relevance and computational viability. Consistent with the journal’s tradition, innovation in optimization methodology and genuine engineering applicability are regarded as equally essential; submissions are expected to articulate clearly how learning contributes to the optimization process and to validate the resulting methods against meaningful baselines.

This Special Issue is organized in conjunction with the Asia Pacific Industrial Engineering and Management Systems (APIEMS) 2026 Conference. Selected high-quality papers presented at the conference will be invited to submit extended versions for possible publication in the Special Issue. The Special Issue is equally open to general submissions from researchers worldwide, and all manuscripts will undergo the standard single-blind peer-review process of Engineering Optimization.

Applications are deliberately left open across the engineering and operational domains in which complex decision-making arises. By bringing together methodological advances and diverse applications under a common theme, the Special Issue aims to clarify how learning can be harnessed to extend the reach and power of optimization in solving complex decision-making problems.

Topic Areas

• Learning-guided and adaptive metaheuristics
• Decision-focused learning and predict-then-optimize approaches
• Reinforcement learning integrated with mathematical programming and combinatorial optimization
• Surrogate-assisted optimization for computationally expensive problems
• Data-driven robust, stochastic, and distributionally robust optimization
• Learning-enhanced multi-objective optimization and multi-criteria decision-making
• Theoretical foundations of learning-augmented optimization
• Production scheduling and planning
• Supply chain and logistics network design
• Vehicle routing and transportation optimization
• Energy systems planning and operation
• Structural and engineering design optimization
• Maintenance, reliability, and resilience optimization
• Resource allocation in service and healthcare systems