 |
International Society of Science and Applied Technologies |
| Optimization of CNN Parameters for Brain Cancer Data Classification | ||||
| Author | Yu-Chi Li
|
|||
| Co-Author(s) | Chien-Wei Wu
|
|||
| Abstract | Brain tumors can generally be categorized into benign and malignant types, with benign tumors such as meningiomas, and malignant tumors such as gliomas. This study will propose an approach to assist physicians in diagnosing four conditions: pituitary gland issues, meningiomas, gliomas, or cases without tumors, aiming to reduce misdiagnosis and enable early treatment. This study aims to explore the application of the Taguchi method to identify the optimal parameter combination for Convolutional Neural Network (CNN) models, reducing experimental time and achieving high accuracy in analyzing brain cancer data. Firstly, we introduce the basic principles and applications of the Taguchi method, emphasizing its significance in quality engineering. Next, utilizing a CNN model to process this dataset, we apply the Taguchi method to optimize parameters such as learning rate, optimizer, and pooling methods, with the goal of improving accuracy.
|
|||
| Keywords | Brain tumors, Taguchi method, Convolutional Neural Network | |||
| Article #: RQD2024-197 | ||||
Proceedings of 29th ISSAT International Conference on Reliability & Quality in Design |