In SAS DLPy, a data scientist trains an image classifier and wants to augment the training data to improve generalization. Which DLPy class is used to specify image augmentation transformations?
Select an answer to reveal the explanation.
Short Explanation
In SAS DLPy, you set up image augmentation through the ImageTable loading pipeline and pass augmentation parameters when fitting the model, so transformations like flipping, rotation, and cropping are applied during training.
Full Explanation
The ImageTable class in DLPy with augmentation parameters passed to model.fit() is correct because image augmentation is configured through the ImageTable (or CASTable) loading process, where parameters such as random flipping, rotation, and cropping are specified as arguments to the data loading or fitting functions. The model.fit() method accepts augmentation configurations through its data handling pipeline. The dlpy.data_augmentation.ImageAugmentation and dlpy.utils.DataAugmentation approaches are wrong because those class names do not exist in standard SAS DLPy. The PROC HPNEURAL with AUGMENT=YES approach is wrong because PROC HPNEURAL does not support image augmentation.