From the course: Neural Networks and Convolutional Neural Networks Essential Training

Unlock the full course today

Join today to access over 22,600 courses taught by industry experts or purchase this course individually.

Preprocessing and loading of data

Preprocessing and loading of data

From the course: Neural Networks and Convolutional Neural Networks Essential Training

Start my 1-month free trial

Preprocessing and loading of data

“

- [Instructor] So we covered the MNIST data set in the neural networks section, and much of this will be repitition. As before, we need to load the libraries first. In neural networks, we only had the fully connected layer, otherwise known as the dense layer. With convolution neural networks, we have far more operations, such as the convolution operation, max pooling, flattening, and also a fully connected or dense layer. We'll use sequential under models, because this will give us a linear stack of neural network layers, and we'll use the MNIST data set as in the previous example, as this is one of the data sets available with Keras. And finally, two categorical allows us to reshape the data and then show the labeled data has 10 categories or bins. Remember that we have absolutely no idea what format the MNIST data is stored in, so we need to read the documentation. And this tells us that a Tupple of Numpy arrays has returned. 60,000 of the images are to be used for training model…

Contents