|
Adaptive Deep Neural Network for Solving Multiclass Problems |
|---|---|
| รหัสดีโอไอ | |
| Creator | Komgrich Onprasonk |
| Title | Adaptive Deep Neural Network for Solving Multiclass Problems |
| Contributor | Tammanoon Panyatip, Panatda Phothinam |
| Publisher | Faculty of Informatics, Mahasarakham University |
| Publication Year | 2569 |
| Journal Title | Journal of Applied Informatics and Technology |
| Journal Vol. | 8 |
| Journal No. | 1 |
| Page no. | 259124 |
| Keyword | Deep Neural Network, Machine Learning, Multiclass, Neural Network |
| URL Website | https://ph01.tci-thaijo.org/index.php/jait |
| Website title | Journal of Applied Informatics and Technology |
| ISSN | 3088-1803 |
| Abstract | In recent years, multiclass classification has gained significant attentiondue to its wide-ranging applications in fields such as healthcare, finance,and image recognition. The ability to accurately classify data into mul-tiple categories is essential for developing intelligent and robust systems.This research compares the performance of several machine learning anddeep learning algorithms for multiclass classification tasks, with a focuson adaptive techniques in neural networks. The evaluated algorithmsinclude Support Vector Machines (SVM), One-vs-Rest Logistic Regres-sion (OvR-LR), Deep Neural Networks (DNN), Dropout-enhanced DNN,and Adaptive Regularization-based DNN. The experimental evaluationwas conducted using both the train–test split and 5-fold cross-validationmethods to ensure result reliability and generalizability. The AdaptiveRegularization-DNN model achieved the highest performance among alltested approaches, with an accuracy of 98.75% under the train–test splitand 97.3% under cross-validation. These results highlight the model’srobustness and its effectiveness in minimizing overfitting in structuredmulticlass classification problems. Performance metrics including preci-sion, recall, F1-score, and accuracy were used to provide a comprehensiveevaluation of each model’s capabilities. |
