Training a Feed-Forward Neural Network Using Particle Swarm Optimizer with Autonomous Groups for Sonar Target Classification

Abstract

Feed-Forward Neural Networks (FFNNs), as one of the wide-spreading Arti¯cial NNs, has been used to solve many practical problems such as classi¯cation of the sonar dataset. Improper selection of the training method, which is an important part of the design process, results in slow convergence rate, entrapment in local minima, and sensitivity to initial conditions. To overcome these issues, the recently proposed method known as \Particle Swarm Optimizer with Autonomous Groups (AGPSO)" has been used in this paper. It is known that the FNNs are very sensitive to the problem's dimension, so clearly applying it to a dataset with large dimension results in poor performance. However, the combination of FNN and AGPSO solves this problem because of the ability of AGPSO in the optimization of high dimension problems. To evaluate the performance of the proposed classi¯er, it is applied to other datasets and the results are compared to the standard PSO, modi¯ed PSO (IPSO), Gravitational Search Algorithm (GSA) and Gray Wolf Optimizer (GWO). Simulation results show that the AGPSO classi¯er provides better performance in terms of convergence speed, entrapment in local minima and classi¯cation accuracy compared to the other algorithms.

An LM-BP Neural Network Approach to Estimate Monthly-Mean Daily Global Solar Radiation Using MODIS Atmospheric Products

“…where i represents the index, xi is one component of the input vector, wi is the weight for each xi, g(•) is the transfer function, which can take many function forms, such as log-sigmoid, tan-sigmoid, and purelin. On the right panel of Figure 2, a BP neural network with three layers containing one input layer, single hidden layer, and one output layer were illustrated; it is a backward propagation of errors [37], the data were propagated from the input layer to the output layer through the hidden layer; while the error was transmitted in the opposite direction, thereby correcting the connection weight (w nm ) of the network. The final error became smaller and smaller.…”

AOAAO: The Hybrid Algorithm of Arithmetic Optimization Algorithm With Aquila Optimizer

“…In addition, there are many improved methods for the shortcomings of different algorithms. Mosavi and Khishe [50] proposed an autonomous groups particles swarm optimization (AGPSO) algorithm. He believes that the behavior of groups is usually determined by the individual or representative that they follow.…”

A novel hybrid predictive maintenance model based on clustering, smote and multi-layer perceptron neural network optimised with grey wolf algorithm

“…Consequently, this has led to the development of new and advanced meta-heuristic algorithms for training MLP such as the hybrid PSO-GSA [42], PSO with Autonomous Groups (PSOAG) [43], Invasive Weed Optimiser (IWO) [44], Chemical Reaction Optimiser (CRO) [45], Stochastic Fractal Search (SFS) [46], Biogeography-Based Optimizer (BBO) [47], Adaptive Best-Mass GSA (ABMGSA) [48], Chimp Optimisation Algorithm (COA) [49], Dragonfly Optimisation Algorithm (DOA) [50], Salp Swarm Optimiser (SSO) [51], Social Spider Optimisation Algorithm (SSOA), Grey Wolf Optimisation (GWO) [41], Equilibrium Optimiser (EO) [52], Sine Cosine Algorithm (SCA) [53], Modified Sine Cosine Algorithm (MSCA) [54], Whale Optimisation Algorithm (WOA) [55], Improved WOA [56], Modified WOA [57] among others.…”

“…The comparison was done by applying the GWO-MLP classifier on both the original dataset (without clustering) and the balanced clustered dataset. The performance of GWO-MLP on these distinct datasets was assessed using the eight performance indicators of classifiers (Eqs (40)(41)(42)(43)(44)(45)(46)(47)…”