Improved PSO based clustering fusion algorithm for multimedia image data projection

作者：Pan, Feng^*; Chen, Deqiang; Lu, Lu

来源：MULTIMEDIA TOOLS AND APPLICATIONS, 2020, 79(13-14): 9509-9522.

DOI：10.1007/s11042-019-08015-z

• Multimedia image data projection
• Clustering fusion
• Inertia weight linear decreasing PSO algorithm
• Optimized base clustering subset
• Global optimization
• Error back propagation

Aiming at the problem that the existing multimedia image clustering fusion algorithm has poor effect on the projection processing of multimedia image data, and the result of the fusion is dispersive, a multimedia image data projection clustering fusion optimization algorithm based on improved particle swarm optimization is proposed. Firstly, using the gradient descent training of error back propagation, the cluster members of the multimedia image data projection are selected to provide an accurate data basis for subsequent cluster fusion. Secondly, each multimedia image data projection base clustering algorithm is selected into the optimized base. Probability of subsets of classes; finally, the improved inertia weight linear decrement PSO algorithm is used for global optimization, and the optimization of multimedia image data projection clustering algorithm is realized. Through experimental verification and analysis, the results show that the algorithm proposed in this paper has high accuracy of projection and clustering of multimedia image data on Data-Set virtual dataset or Iris actual dataset, and the average accuracy is above 90%.