| Autor: |
Nabilah Hanani Hider, Anis Salwa Binti Mohd Khairuddin, Effariza Binti Hanafi |
| Rok vydání: |
2022 |
| Zdroj: |
Proceedings of International Technical Postgraduate Conference 2022. |
| ISSN: |
2582-3922 |
| DOI: |
10.21467/proceedings.141.9 |
| Popis: |
Lung cancer is one of the most common cancers worldwide that leads to small survival rate. It is important to detect the presence of these harmful cells in human body at early stages to prevent it from worsening. The primary goal of this study is to propose an efficient lung cancer image classification model using deep learning method. The cancer image classification framework is proposed by using transfer learning with Convolutional Neural Network (CNN) to classify three categories of 5,100 cancer images namely lung adenocarcinoma, lung squamous cell carcinoma and benign lung tissues obtained from the dataset. Several experiments have been performed to improve the VGG19 model performance by varying the optimizers including RMSprop, Adam and SGD. The performance of all experiments conducted were analyzed based on the training and validation curves, classification reports and the confusion metrics. |
| Databáze: |
OpenAIRE |
| Externí odkaz: |
|
