What is a Feedforward Neural Network?
A feedforward neural network (FNN) is a type of artificial neural network where connections between the nodes do not form cycles. Information moves in one direction—from input nodes, through hidden layers (if any), to the output nodes.
Key Characteristics
- Layers are arranged sequentially.
- No recurrent connections.
- Each neuron applies a weighted sum followed by an activation function.
Simple Example: 2‑2‑1 Network
This network has two inputs, two hidden neurons, and one output neuron. Use the form below to see the forward pass in action.
Mathematical Formulation
h₁ = σ(w₁₁·x₁ + w₁₂·x₂ + b₁)
h₂ = σ(w₂₁·x₁ + w₂₂·x₂ + b₂)
y = σ(v₁·h₁ + v₂·h₂ + bₒ)σ denotes the activation function, commonly the sigmoid, ReLU, or tanh. In the interactive demo above we use the sigmoid function:
σ(z) = 1 / (1 + e⁻ᶻ)