Phase 4 · Deep learning · Neural networks
The artificial neuron
Weighted sum plus a nonlinearity.
Why it matters
Every neural network, from a spam filter to a model with hundreds of billions of parameters, is built from one small unit repeated over and over. You've already trained one without calling it that: logistic regression. Understanding exactly what a single neuron can and can't do is the cleanest way to see why networks need layers, which is the story of this phase.
What you'll be able to do
- Compute a neuron's output: a weighted sum plus a bias, passed through an activation
- Recognize logistic regression as a single neuron
- See a neuron's decision boundary as a straight line
- Explain why one neuron can't learn XOR, and why that matters
Before this lesson
Next in this phase: Activation functions