Apache Spark Deep Learning Cookbook

Getting ready

The following key points need to be considered for this recipe:

• The MIT-CBCL dataset is composed of 3,240 images (324 images per subject). In our model, we will make arrangements to augment the data in order to increase model robustness. We will employ techniques such as shifting the subject, rotation, zooming, and shearing of the subject to obtain this augmented data.
• We will use 20% of the dataset to test our model (648 images) by randomly selecting these images from the dataset. Similarly, we randomly select 80% of the images in the dataset and use this as our training dataset (2,592 images).
• The biggest challenge is cropping the images to the exact same size so that they can be fed into the neural network.
• It is a known ...