A Novel Hybrid Algorithm for Efficient Anomaly Detection Using Machine Learning and Optimization Techniques

EasyChair Preprint 15470

Abstract

Anomaly detection plays a critical role in various fields, including cybersecurity, finance, and
healthcare. Despite advancements in machine learning, the development of robust algorithms that
balance computational efficiency and detection accuracy remains a challenge. This paper introduces
a novel hybrid algorithm combining Particle Swarm Optimization (PSO) with a Neural Network (NN)
to enhance anomaly detection. The proposed method leverages PSO for feature selection and
hyperparameter optimization, while the NN ensures robust classification. Experimental results on
benchmark datasets demonstrate significant improvements in accuracy and computational
performance compared to existing approaches.

Keyphrases: Particle Swarm Optimization, anomaly detection, hybrid algorithm, machine learning, neural networks

@booklet{EasyChair:15470,
  author    = {James Rajez and Mo Zhang and Mehmmet Amin},
  title     = {A Novel Hybrid Algorithm for Efficient Anomaly Detection Using Machine Learning and Optimization Techniques},
  howpublished = {EasyChair Preprint 15470},
  year      = {EasyChair, 2024}}