A data scientist uses SAS DLPy to build an image classification model. Which deep learning architecture is most appropriate for classifying images of product defects?
Select an answer to reveal the explanation.
Short Explanation
You want the convolutional neural network using ResNet because it learns edges, textures, and shapes from images, which is exactly what product-defect classification needs. RNN/LSTM models are built for sequences, not spatial image patterns.
Full Explanation
The convolutional neural network using ResNet architecture is the appropriate choice because CNNs exploit spatial hierarchies in image data through convolutional layers that detect edges, textures, and shapes, making them the standard approach for image classification tasks including defect detection. The recurrent neural network with LSTM cells is designed for sequential data like time series or text, so it is not the natural fit for static product-defect images. The feedforward neural network with only fully connected layers ignores spatial structure and is impractical for high-resolution images. The transformer model with self-attention on raw pixel values is computationally prohibitive without convolutional preprocessing and is not a standard SAS DLPy workflow.