Detecting fake accounts on Instagram using machine learning and hybrid optimization algorithms

Abstract

In this thesis, we propose a hybrid method for detecting fake accounts on Instagram by using the Binary Grey Wolf Optimization Algorithm (BGWO) and Particle Swarm Optimization (PSO). By combining these two algorithms, we aim to leverage their complementary strengths and enhance the overall optimization performance. We evaluate the proposed hybrid method by using four classifiers, Support Vector Machine (SVM), Logistic Regression (LR), K-Nearest Neighbor (KNN), and Artificial Neural Network (ANN). The dataset used for the experiments contains 65329 Instagram accounts. We extract features from each account, including profile information, posting behavior, and engagement metrics by feature selection using PSO. The results show that the Hybrid optimization method (BGWOPSO) significantly outperformed both Binary Grey Wolf Optimization (BGWO) and Particle Swarm Optimization (PSO) methods when using several performance measures including accuracy, precision, recall, and AUC selecting the best optimal features.