Training Feedforward Neural Network Using Enhanced Black Hole Algorithm: A Case Study on COVID-19 Related ACE2 Gene Expression Classification

Abstract

The aim of this paper is twofold. First, black hole algorithm (BHA) is proposed as a new training algorithm for feedforward neural networks (FNNs), since most traditional and metaheuristic algorithms for training FNNs suffer from the problem of slow coverage and getting stuck at local optima. BHA provides a reliable alternative to address these drawbacks. Second, complementary learning components and Levy flight random walk are introduced into BHA to result in a novel optimization algorithm (BHACRW) for the purpose of improving the FNNs' accuracy by finding optimal weights and biases. Four benchmark functions are first used to evaluate BHACRW's performance in numerical optimization problems. Later, the classification performance of the suggested models, using BHA and BHACRW for training FNN, is evaluated against seven various benchmark datasets: iris, wine, blood, liver disorders, seeds, Statlog (Heart), balance scale. Experimental result demonstrates that the BHACRW performs better in terms of mean square error (MSE) and accuracy of training FNN, compared to standard BHA and eight well-known metaheuristic algorithms: whale optimization algorithm (WOA), biogeography-based optimizer (BBO), gravitational search algorithm (GSA), genetic algorithm (GA), cuckoo search (CS), multiverse optimizer (MVO), symbiotic organisms search (SOS), and particle swarm optimization (PSO). Moreover, we examined the classification performance of the suggested approach on the angiotensin-converting enzyme 2 (ACE2) gene expression as a coronavirus receptor, which has been overexpressed in human rhinovirus-infected nasal tissue. Results demonstrate that BHACRW-FNN achieves the highest accuracy on the dataset compared to other classifiers.

An Efficient Binary Sand Cat Swarm Optimization for Feature Selection in High-Dimensional Biomedical Data

“…Various techniques, including mutation [5], Lévy flight [19], and opposition-based learning (OBL) [20], have been used in the literature to increase NIOA's exploration capabilities. OBL broadens the search range by computing the inverse of the existing viable solution and locating candidate solutions in more ideal places.…”

MSXFGP: combining improved sparrow search algorithm with XGBoost for enhanced genomic prediction

“…For example, Elnaz Pashaei [ 21 ] combined the swarm intelligence algorithm COOT with the metaheuristic algorithm Simulated Annealing for feature selection in high-dimensional microarray data, enhancing the experimental outcomes. Elham Pashaei [ 22 ] employed an improved Black Hole Algorithm (BHA) to find the optimal weights and biases for Feedforward Neural Networks (FNN), increasing the model’s accuracy. Moreover, some researchers have tried to use optimization algorithms for simultaneous feature selection and parameter optimization to enhance model results.…”

Hybrid Hypercube Optimization Search Algorithm and Multilayer Perceptron Neural Network for Medical Data Classification

“…Optimization algorithms (OAs) are offered as viable alternatives to gradient-based MLP training approaches in this regard. Several works have been published in the literature, including group search optimizer (GSO) [5], symbiotic organisms search (SOS) [6] algorithm, lightning search algorithm (LSA) [7], ant lion optimizer (ALO) [8], Krill herd algorithm (KHA) [9], grasshopper optimization algorithm (GOA) [10,11], artificial bee colony (ABC) [12], social spider optimization algorithm (SSO) [13], hybrid of ABC and dragonfly algorithm (DA) [14], artificial ant colony optimization (ACO) [15], particle swarm optimization (PSO) [16], cuckoo search (CS) [17,18], moth-flame optimization (MFO) [19,20], whale optimization algorithm (WOA) [21], gray wolf optimizer (GWO) [22,23], black hole algorithm (BHA) [24], invasive weed optimization [25], multiverse optimizer algorithm (MOA) [26,27], bat algorithm [28], and salp swarm algorithm (SSA) [29].…”