After the No Free Lunch Theorems (NFLT), twenty years ago, the necessity of designing algorithms which specifically address some problems was expressed. As a result, the approach of designing algorithm on the basis of the most various inspirations (genetics, behavior of birds, fish, etc.) and with general purpose has been replaced by a design especially tailored to solve the problems. This modern and rapidly growing interpretation of optimization problems and algorithms is referred to with several notations according to some design differences and philosophy underneath. We will use the comprehensive term “Metaheuristic Optimization.”

Modern Metaheuristic Optimization approaches can be divided into two main categories: algorithms that perform change during the run on the basis of the success of the components thus adapting to the problem and algorithms that are designed after a thorough problem examination. These two approaches can be applied in both algorithm design studies and real-world applications.

This special issue focuses on novel modern Metaheuristic Optimization algorithms, that is, algorithmic approaches that, in accordance with the NFLT, integrate the problem and its features within the optimization process. More specifically, the articles in this special issue will focus on the design of the metaheuristics that take into account the problem features; thus the problem algorithm appears in some way in the search algorithm that tackles it either at design time or at run time. Opposite to the common practice of designing algorithms by randomly combining existing algorithms or perturbing and complicating successful algorithmic framework, we are interested in investing those approaches where each algorithmic component is thoughtfully used to tackle the challenges posed by the specific problem.

Prospective authors from academia and industry are invited to submit their original and unpublished contributions to this special issue. Studies focusing on novel algorithm design approaches and attractive real-world applications are both welcome.

Potential topics include but are not limited to the following:

• Hyperheuristic algorithms
• Memetic computing
• Adaptive and self-adaptive schemes
• Membrane computing algorithms
• Fitness landscape analysis
• Problem examination techniques
• Studies on problem features, such as separability, ill-conditioning, and multimodality
• Theoretical studies about algorithm functioning
• Domain specific implementation