Machine Learning vs Deep Learning

Deep learning is a subset of machine learning, which is a subset of AI.

Point	Machine learning	Deep learning
Features	Hand-crafted by experts (texture, shape, intensity)	Learned automatically from raw images
Models	SVM, random forest, k-NN, logistic regression	Multi-layer neural networks: CNN, U-Net, transformers
Data needed	Small to medium	Very large
Computing	CPU is enough	GPU required
Training time	Short	Long
Interpretability	Easier	Black box
Accuracy on images	Good	Usually higher

Medical examples

• Machine learning: radiomics features extracted from a tumour CT, then a random forest or SVM classifies benign vs malignant; predicting disease risk from lab and patient data.
• Deep learning: a CNN detects pneumonia on chest X-ray; U-Net segments brain tumours on MRI; deep networks find diabetic retinopathy on retinal images; AI reconstructs low-dose CT.

Exam tip

Draw the nested diagram and write the feature-extraction difference first.