TensorFlow Machine Learning Projects

Pre-processing data and defining the model

We need to pre-process our data for making it ready to feed into our model, define our model, and create an evaluation metric:

1. Pre-process the data by ensuring the images are of shape 28x28x1 and converting the pixels into a float type variable for training. Also, here we define NUM_CLASSES = 10 as there are 10 different digits in the images.

x_train = x_train.reshape(x_train.shape[0], IMAGE_SIZE, IMAGE_SIZE, 1)x_test = x_test.reshape(x_test.shape[0], IMAGE_SIZE, IMAGE_SIZE, 1)x_train = x_train.astype('float32')x_test = x_test.astype('float32')Next, we normalize the image pixels by 255 as follows:x_train /= 255x_test /= 255And finally, we convert the class labels to one hot for training as follows: ...