Neural Networks
AI · Intermediate
Definition
A neural network is layers of connected neurons that multiply inputs by weights, add a bias and apply an activation.
Neural networks power image recognition, speech and language models.
Diagram
Flowchart
Example: One neuron
x = [1.0, 2.0]; w = [0.5, -0.25]; b = 0.1
z = sum(a*b_ for a, b_ in zip(x, w)) + b
print(max(0, z))Output: 0.1
Key points
- Activations add non-linearity
- Backpropagation computes gradients
- Deep = many hidden layers
Interview questions
Why do we need activation functions?
Without them, stacked layers collapse into one linear function.