Question

A data science team is building an image classifier to detect defective parts from thousands of high-resolution factory photos. A traditional machine learning approach would require engineers to manually design features (edges, textures, shapes) before training a classifier. The team instead chooses a deep learning convolutional neural network. What is the primary advantage that motivates this choice?

Accepted Answer

Deep neural networks automatically learn hierarchical feature representations directly from raw pixel data, removing the need for manual feature engineering.

Answer

Deep learning models always require far less training data than classical machine learning methods.

Answer

Convolutional neural networks eliminate the need for any labeled training examples.

Answer

Deep learning models can be trained effectively on CPUs alone without needing GPU acceleration.