Building a model to automatically classify different types of brain tumors based on the images
The result shows that the model has an accuracy rate of 0.97 and loss rate 0.06. It has a very good performance for predicting the images
I used Neural Network by utilizing Transfer Learning. The Image Augmentation process was also carried out before the model training stage
OVERVIEW
This is my personal project provided by Coursera to build a model that can automatically classify the different types of brain tumors based on the existing images. The dataset has 3,285 total images divided into 4 classes (Meningioma Tumor, Pituitary Tumor, Glioma Tumor, and no tumor).
For the initial stage, I cropped the images so that the prediction results were more accurate and faster. Then, I did an image Augmentation which divided the image into training, test, and validation subdirectories with a validation split value of 0.2.
I used transfer learning at the architectural model compiling and model training stage . Transfer learning is a method of transferring and using the data of a pre-trained model in advance so there is no need to train the model from scratch.
Based on the results of model training, it has an accuracy rate of 0.97 and loss rate 0.06. It has a very good performance for predicting the images
