Big Data Value Calculation Method Based on Particle Swarm Optimization Algorithm.

Area of Science:

• Artificial Intelligence
• Computational Intelligence
• Data Science

Background:

• Social Insect (SI) inspired systems utilize collective behavior in decentralized networks.
• Big data presents challenges for traditional computer techniques, particularly in data categorization.
• Existing methods struggle with learning from large datasets and optimizing computational efficiency.

Purpose of the Study:

• To propose a novel CIN-big data value calculation algorithm, Big Data-based Particle Swarm Optimization (BD-PSO).
• To improve the operating efficiency and convergence speed of Particle Swarm Optimization (PSO) in big data environments.
• To reduce calculation time and enhance the accuracy of big data classification.

Main Methods:

• Development of the Big Data-based Particle Swarm Optimization (BD-PSO) algorithm.
• Application of BD-PSO to big data categorization tasks.
• Performance evaluation using four UCI benchmark datasets (wine, iris, blood transfusion, zoo).
• Comparison with existing methods like Support Vector Machines (SVM) and CG-CNB.

Main Results:

• The BD-PSO algorithm demonstrated improved convergence speed and computational efficiency.
• Achieved 92% accuracy, 92% precision, and 92% recall on benchmark datasets.
• Execution time was reduced to 149 ms, significantly outperforming existing approaches.
• BD-PSO provided robust solutions for optimization problems in big data classification.

Conclusions:

• The BD-PSO algorithm is an effective intelligent technique for big data classification and optimization.
• BD-PSO significantly enhances computational efficiency and accuracy compared to traditional methods.
• The approach offers a robust solution for handling large datasets and complex classification tasks.