Hannah I want to understand what dropout means in deep learning models. How does it help prevent overfitting during neural network training? Can someone also explain how dropout layers work with simple examples?