CNN Optimization with Adam Optimizer for Mushroom Data Classification

Abstract

Mushroom identification is an important task in food safety and biological classification due to the morphological similarities between edible and poisonous species. Traditional identification methods often rely on expert knowledge and manual observation, which may be time-consuming and prone to human error. Therefore, this study proposes an optimized Convolutional Neural Network (CNN) model using Adam Optimizer for mushroom classification. The proposed approach employs image preprocessing techniques including resizing, normalization, and data augmentation to improve data quality and model robustness. The dataset was divided into training, validation, and testing subsets to ensure reliable model evaluation. CNN architecture consisting of convolutional, pooling, and fully connected layers was implemented and trained using Adam Optimizer with adaptive learning rate mechanisms. Experimental results demonstrated that the proposed model achieved high classification performance, with training accuracy reaching 98.21% and validation accuracy of 96.54%, accompanied by consistently decreasing loss values. Confusion matrix analysis further confirmed reliable classification capability with minimal misclassification between mushroom classes. Comparative evaluation indicated that Adam Optimizer outperformed conventional optimization methods in terms of convergence speed and classification accuracy. These findings suggest that CNN optimized with Adam provides an effective and reliable framework for automated mushroom classification and has strong potential for applications in food safety, agriculture, and intelligent biological image recognition systems.