Parallel stacked autoencoder with particle swarm optimization for community detection in complex networks

A deep memory bare-bones particle swarm optimization algorithm for single-objective optimization problems

A deep memory bare-bones particle swarm optimization algorithm (DMBBPSO) for single-objective optimization problems is proposed in this paper. The DMBBPSO is able to perform high-precision local search while maintaining a large global search, thus providing a reliable solution to high-dimensional complex optimization problems. Normally, maintaining high accuracy while conducting global searches is an important challenge for single-objective optimizers. Traditional particle swarms optimizers can rapidly lose the diversity during iterations and are unable to perform global searches efficiently, and thus are more likely to be trapped by local optima. To address this problem, the DMBBPSO combines multiple memory storage mechanism (MMSM) and a layer-by-layer activation strategy (LAS). The MMSM catalyzes a set of deep memories to increase the diversity of the particle swarm. For every single particle, both of the personal best position and deep memories will be used in the evaluation process. The LAS enables the particle swarm to avoid premature convergence while enhancing local search capabilities. The collaboration between MMSM and LAS enhances the diversity of the particle swarm, which in turn enhances the robustness of the DMBBPSO. To investigate the optimization ability of the DMBBPSO for single-objective optimization problems, The CEC2017 benchmark functions are used in experiments. Five state-of-the-art evolutionary algorithms are used in the control group. Finally, experimental results demonstrate that the DMBBPSO can provide high precision results for single-objective optimization problems.

Tuning Machine Learning Models Using a Group Search Firefly Algorithm for Credit Card Fraud Detection

Recent advances in online payment technologies combined with the impact of the COVID-19 global pandemic has led to a significant escalation in the number of online transactions and credit card payments being executed every day. Naturally, there has also been an escalation in credit card frauds, which is having a significant impact on the banking institutions, corporations that issue credit cards, and finally, the vendors and merchants. Consequently, there is an urgent need to implement and establish proper mechanisms that can secure the integrity of online card transactions. The research presented in this paper proposes a hybrid machine learning and swarm metaheuristic approach to address the challenge of credit card fraud detection. The novel, enhanced firefly algorithm, named group search firefly algorithm, was devised and then used to a tune support vector machine, an extreme learning machine, and extreme gradient-boosting machine learning models. Boosted models were tested on the real-world credit card fraud detection dataset, gathered from the transactions of the European credit card users. The original dataset is highly imbalanced; to further analyze the performance of tuned machine learning models, in the second experiment performed for the purpose of this research, the dataset has been expanded by utilizing the synthetic minority over-sampling approach. The performance of the proposed group search firefly metaheuristic was compared with other recent state-of-the-art approaches. Standard machine learning performance indicators have been used for the evaluation, such as the accuracy of the classifier, recall, precision, and area under the curve. The experimental findings clearly demonstrate that the models tuned by the proposed algorithm obtained superior results in comparison to other models hybridized with competitor metaheuristics.