C27 1 hr 30 min
Machine learning, deep learning, and expert systems
The basic concept of machine learning and how it differs from expert systems.
Machine learning
Machine learning builds models from data rather than explicit rules. Common forms:
- Supervised — learn from labelled examples.
- Unsupervised — find structure in unlabelled data.
- Reinforcement — learn by reward and penalty.
Deep learning
Deep learning is machine learning with many-layered neural networks. It excels when data is large and unstructured, such as images or audio, but needs more data and compute than classical methods.
Expert systems
An expert system captures human expertise as rules and an inference engine. It is explainable and predictable but brittle outside its rule set.
| Approach | Source of knowledge | Strength |
|---|---|---|
| Expert system | Hand-written rules | Explainable |
| Machine learning | Data | Adapts to patterns |
| Deep learning | Large data + neural nets | Handles complex input |
Check your understanding
- When would an expert system be preferable to machine learning?
- What makes deep learning different from classical machine learning?