Overview Swarm-based metaheuristics in automatic programming: a survey Juan L. Olmo, 1 José R. Romero 1 and Sebastián Ventura 1,2∗ On the one hand, swarm intelligence (SI) is an emerging ﬁeld of artiﬁcial intelli- gence that takes inspiration in the collective and social behavior of different groups of simple agents. On the other hand, the automatic evolution of programs is an active research area that has attracted a lot of interest and has been mostly pro- moted by the genetic programming paradigm. The main objective is to ﬁnd com- puter programs from a high-level problem statement of what needs to be done, without needing to know the structure of the solution beforehand. This paper looks at the intersection between SI and automatic programming, providing a survey on the state-of-the-art of the automatic programming algorithms that use an SI meta- heuristic as the search technique. The expression of swarm programming (SP) has been coined to cover swarm-based automatic programming proposals, since they have been published to date in a disorganized manner. Open issues for future research are listed. Although it is a very recent area, we hope that this work will stimulate the interest of the research community in the development of new SP metaheuristics, algorithms, and applications. © 2014 John Wiley & Sons, Ltd. How to cite this article: WIREs Data Mining Knowl Discov 2014. doi: 10.1002/widm.1138 INTRODUCTION B io-inspired algorithms 1 are a kind of algorithm based on biological systems that mimic the prop- erties of these systems in nature. They are attractive from a computational point of view due to their broad application areas and their simplicity and random components, inherited from natural systems. Most bio-inspired algorithms are easy to implement and their complexity is relatively low, and these algo- rithms, though simple, can search multimodal land- scape with sufﬁcient diversity and ability to escape any local optimum. 2 Bio-inspired computing is an active and promising research ﬁeld in algorithm design, and includes paradigms such as artiﬁcial neural networks (ANNs), 3 evolutionary algorithms (EAs), 4 artiﬁcial immune systems (AIS), 5 and swarm intelligence (SI), 6 among others. ∗ Correspondence to: sventura@uco.es 1 Department of Computer Science and Numerical Analysis, Univer- sity of Córdoba, Córdoba, Spain 2 Department of Computer Science, King Abdulaziz University, Jeddah, Saudi Arabia Conﬂict of interest: The authors have declared no conﬂicts of interest for this article. In particular, SI focuses on the development of multi-agent systems inspired by the collective behav- ior of simple agents. The general objectives of the swarm are pursued by means of individuals’ indepen- dent actions, which can interact locally both with one another and with the environment. A global, intel- ligent and coordinated behavior emerges from these independent actions. 7 Representative types of SI are particle swarm optimization (PSO), 8 which deals with the movement of ﬂocks of birds or schools of ﬁshes; ant colony optimization (ACO), 9 which takes inspira- tion from the behavior and self-organization capabili- ties of ant colonies; or bee swarm intelligence (BSI), 10 which models some of the features of honey bees. SI algorithms are expanding and becoming increasingly popular in many disciplines and applications, mainly because of their ﬂexibility and efﬁciency in solving a wide range of high complex problems. 11 On the other hand, automatic programming is an active research ﬁeld with applications in many domains that has become very popular mainly due to the widespread use of the GP paradigm. 12 Automatic programming is a method that uses a search technique to automatically construct a computer program that solves a given problem, without requiring the user © 2014 John Wiley & Sons, Ltd.