Regional Economic Development Level using Long Short-Term Memory based Perturbation Zebra Optimization Algorithm

The regional economic development level defines that industrialization, economic growth, prosperity of particular geographical area and infrastructure. It is inclined by factors like employment, investigation, policy framework, and technological advancement. Determining this level assist in strategic planning in sustainable growth. However, development level faces challenges because of disparities in industrial growth, resource allocation, and policy efficiency that results in imbalanced progress. This research proposes, Long Short-Term Memory based Perturbation Zebra Optimization Algorithm (LSTM-PZOA) to develop the regional economic level effectively. In economic indicators, LSTM captures long-term dependencies which enable accurate trend for regional economic development whereas PZOA is applied to fine-tune the LSTM parameter values. In traditional ZOA, the Particle Swarm Perturbation Mechanism (PSPM) is incorporated to explore the search space effectively by solving local optima issue. Compared to existing methods like Deep Neural Network (DNN) and Recurrent Neural Network (RNN), the proposed LSTM-PZOA achieves a high accuracy and f1-score of 87.32% and 82.77% respectively.