Asked by ABAINEH

What characteristic does deep learning have, that traditional machine learning does not?

Answers

There are no human answers yet.
Answered by GPT-5 mini
Deep learning learns hierarchical, multi-layer representations of the data (automatic feature learning) from raw inputs — i.e., depth.

Why that matters:
- Deep networks automatically extract low‑level to high‑level features across many nonlinear layers (edges → textures → object parts → objects in vision), so they can be trained end-to-end from raw data.
- Traditional ML typically relies on manually engineered features or shallow transformations plus a classifier (SVM, logistic regression, random forest), rather than learning layered abstractions.

Tradeoffs: deep learning usually needs much more data and compute and is often less interpretable, but its core distinguishing characteristic is that it learns hierarchical representations automatically.