There are countless computer algorithms that simulate biological behaviour from leaping frogs, to bat foraging, from cuckoo search to ant colony optimisation. They all have something in common, the algorithm behaves like a collective intelligence, taking on the call and response of a shoal of fish or a murmuration of starlings, and all those other patterns in nature. Writing in the International Journal of Swarm Intelligence, a team from India discusses the state of the art in a unique algorithm based on a biological system – the spider monkey.
Spider monkeys have a “fusion-fission” social structure where a large social group will split into smaller hordes or vice versa depending on the accessibility and availability of food. Janmenjoy Nayak of Aditya Institute of Technology and Management in Andhra Pradesh, India, and colleagues have looked at the spider monkey optimisation algorithm, which embeds this behaviour to allow it to solve otherwise intractable problems. SMO algorithms are, the team reports, particularly useful in solving electrical and electronic engineering, wireless sensor network, pattern recognition, power system and networks, and data mining problems.
Their survey of the state of the art in SMO and its variants and how it can successfully deal with difficult authentic world optimization problems should serve to inspire practitioners and researchers to innovate in this area even more. Moreover, the success of the SMO hints at the potential of different behaviour in other species such as squirrel monkey, vervet monkey, and proboscis monkey, that might also be simulated to good effect.
Nayak, J., Vakula, K., Dinesh, P. and Naik, B. (2019) ‘Spider monkey optimisation: state of the art and advances’, Int. J. Swarm Intelligence, Vol. 4, No. 2, pp.175–198.