Multi-Column RBF Neural Network Using Adaptive and Non-Adaptive Particle Swarm Optimization
Researchers have developed novel multi-column radial basis function neural network (RBFN) approaches, MC-PSO and MC-APSO, to address scalability challenges with large datasets. These methods leverage particle swarm optimization (PSO) and its adaptive variant (APSO) within a parallel RBFN structure. By training individual RBFNs on spatial subsets of data, the proposed techniques aim to improve accuracy and speed compared to existing gradient-based and swarm-based methods. AI
IMPACT These new RBFN training methods could lead to more efficient and accurate AI models for large-scale datasets.