Large-Scale Global Optimization Repository
There are many factors that contribute to the complexity of optimization problems. One such factor is the number of decision variables involved in an optimization problem. A linear growth in the number of decision variables results in an exponential growth in the size of the search space. This phenomenon which is known as the curse of dimensionality is a major hindrance to the performance of optimization algorithms.
A considerable number of real-world optimization problems exhibit large-scale properties. As we advance in science and technology, the need for solving such large-scale problems will continue to grow. In addition to the large-scale property, some optimization problems are black-box. This means that the algebraic form of the objective function or the derivative information are not available. This property rules out the possibility of using classic gradient-based solvers. Fortunately, numerous derivative-free optimization algorithms have been developed for solving black-box optimization problems.
In recent years, there has been a growing interest in adopting various meta-heuristic algorithms for solving large-scale black-box optimization problems. Development of several large-scale global optimization benchmark suites, and publication of hundreds of scholarly articles in mainstream conferences and journals are a testament to this trend.
The aim of this repository is to collect the relevant publications, and benchmark problems for large-scale global optimization in one central location. This may help the research community in the following major ways:
To facilitate access to the latest developments and the state-of-the-art algorithms in the field of large-scale global optimization.
To provide a platform for conducting comparative studies and tracking the progress of large-scale global optimization algorithms.
To help the researchers in finding the gaps in the field and encouraging the development of new algorithms to fill the perceived gaps.