The Application of Particle Swarm Optimization in Neural Networks

Abstract

The birth of the neural network model has provided a technological platform for the development of various industries. Among them, the branch network scientific research management model represented by the scientific research field is particularly profitable. China’s traditional research management system has been unable to meet the requirements of the present era and the management system reform is imperative. Therefore, based on Particle Swarm Optimization (PSO), this dissertation studies the neural network model based on the prerequisite that the branch network research management system retains the advantages of the traditional model. It tracks the optimal value of current search through cross and mutation. The optimal solution value is finally determined by solving the relationship between particle weight and inertia weight. In the final test, we establish the feasibility of the program.

Sparse Subgraph Prediction Based on Adaptive Attention

“…Based on heuristic methods, strategies using metaheuristics [22] are combined with the Particle Swarm Optimization (PSO) [23] approach and the topological properties of social network graphs to locate missing relationships in symbolic social networks. Ref.…”