Improving Raisin Grains Classification with a Hybrid PSO-NN Approach
Abstract views: 29 / PDF downloads: 113
Keywords:Raisin Grain Classification, Hybrid Approach, PSO-NN, KNN, Random Tree (RT)
– Raisin grain classification is crucial in the food industry for maintaining product quality. Traditional classification techniques can be labor-intensive and time-consuming, presenting significant challenges. To address these issues, this study proposes a hybrid approach for raisin classification that combines Particle Swarm Optimization (PSO) and Artificial Neural Networks (ANN). The PSO algorithm is utilized to optimize ANN models with the aim of enhancing the accuracy of raisin grain classification. Our research, conducted on a dataset consisting of 900 raisin grains from two distinct categories, evaluates the performance of the proposed hybrid PSO-ANN method in comparison to k-Nearest Neighbor (KNN) and Random Tree (RT). The hybrid PSO-NN approach achieved a remarkable classification performance, demonstrating an accuracy rate of 100%, outperforming other methods under evaluation. The respective accuracies of KNN and RT were 87.39% and 94.91%. This outstanding performance underscores the efficacy of integrating PSO optimization with ANN in the field of raisin grain classification. The results suggest that the hybrid PSO-ANN approach surpasses other methods in classification accuracy, indicating its potential to advance raisin grain classification within the food industry.