In this segment ,you will learn to build CNNs in Keras .Specifically, you will learn to build and train a CNN on the MNIST dataset to classify the digits into one of the ten classes (0-9).
This is a text only page (with an IPython notebook) whose objective to make you familiar with building CNNs in Keras. In the nest few segments, the professor will demonstrate some more experiments with CNNs using python + Keras using the CIFAR-10 dataset.
please download the notebook below and go through it. You do not need to download the dataset separately, it can be download from Keras directly (as done in the notebook).
please run thisnotebook locally, not on a GPU.You will use the GPS in the upcoming segments. Make sure you understand the section ‘Understanding Model Summary‘ in the notebook well – it will be required to solve the questions on the next page.
NOTE:The code in all of the notebooks privided for this session is updated occasionally. Therefore it may vary slightly from the one shown in the videos.
In the next segment, you will test your understanding of the concepts covered in the notebook by solving some questions on the VGG-16 architecture.
Report an error