Adaptive Deep Neural Network for Solving Multiclass Problems

In recent years, multiclass classification has gained significant attention due to its wide-ranging applications in fields such as healthcare, finance, and image recognition. The ability to accurately classify data into multiple categories is essential for developing intelligent and robust systems. This research compares the performance of several machine learning and deep learning algorithms for multiclass classification tasks, with a focus on adaptive techniques in neural networks. The evaluated algorithms include Support Vector Machines (SVM), One-vs-Rest Logistic Regression (OvR-LR), Deep Neural Networks (DNN), Dropout-enhanced DNN, and Adaptive Regularization-based DNN. The experimental evaluation was conducted using both the train–test split and 5-fold cross-validation methods to ensure result reliability and generalizability. The Adaptive Regularization-DNN model achieved the highest performance among all tested approaches, with an accuracy of 98.75% under the train–test split and 97.3% under cross-validation. These results highlight the model’s robustness and its effectiveness in minimizing overfitting in structured multiclass classification problems. Performance metrics including precision, recall, F1-score, and accuracy were used to provide a comprehensive evaluation of each model’s capabilities.