Integration of NIC algorithms and ANN: A review of different approaches

Abstract

Nature inspired algorithms are inspired from our environment. These algorithms are inspired by the behaviour of insects, birds and animals etc. In the same way, the artificial neural network (ANN) is the replication of a human brain. Both types of algorithms are used to solve the complex and nonlinear problems of real world. In this paper, we have presented different nature inspired computing (NIC) based techniques, which are used to design and evolve ANN. The results of our study have shown that the manual designing of ANN is very difficult, so we need to design ANN automatically. The Back propagation (BP) method can be used with the combination of any nature inspired algorithm like PSO, ABC or ACO, as individually, they are not giving accurate results. The hybrid algorithms have demonstrated better results as compared to the individual uses of both.