In this project, normal, non-covid pneumonia and covid-19 detection studies will be carried out from chest x-ray pictures. Convolutional Neural Networks will be created and trained for prediction.
16,769 Chest X-Ray images were used for this project. The training and test sets were separated from the data as 90-10. While training, we will allocate 20% of 90% as validation.
At the end of the training:
- Train Performance : %95
- Test Performance : %86
You can use it to access the dataset
or
You can find the summary of the dataset here.