A Comprehensive Overview of PSO-LSTM Approaches: Applications, Analytical Insights, and Future Opportunities

ABSTRACT

The combination of particle swarm optimization algorithm with long short-term memory has led to new horizons in data analysis and solving complex problems. The particle swarm optimization has been used to solve complex and hard optimization problems since 1995 and has gained extraordinary popularity among researchers. The combination of particle swarm optimization-long short-term memory is carried out in order to simultaneously utilize the global search capability and convergence ability of the particle swarm optimization to optimize long short-term memory in modeling long-term temporal dependencies. The goal of the particle swarm optimization-long short-term memory model is to increase the prediction accuracy and reduce the error in solving complex and dynamic real-world problems. In this paper, a comprehensive and structured look at the scientific literature of particle swarm optimization-long short-term memory models between 2019 and June 2025 has been conducted. By categorizing the articles according to the publication date and place of publication, it was found that prominent publishers such as MDPI, Springer, Elsevier, and IEEE played a major role in the publication of particle swarm optimization-long short-term memory models in 2024. This paper aims to provide a clear overview of the potential applications of particle swarm optimization-long short-term memory in various domains. particle swarm optimization-long short-term memory has been widely used in engineering systems design, time series forecasting, and the oil and gas industry. This paper analyses the strengths and weaknesses, as well as the challenges of complexity in hybrid architectures, and issues related to scalability and optimization. Finally, future directions are proposed with an emphasis on performance enhancement and development of adaptable solutions for real-world problems.