Neural Networks
Understand the basic role of an artificial neural network and how a neuron transforms inputs into an output.
Core explanation
An artificial neural network is a model composed of interconnected computational units commonly called neurons. Each neuron receives numerical inputs, combines them using learned weights and a bias, and applies an activation function.
The neuron is a mathematical computation. It is not a tiny human brain and should not be described as independently understanding the input.
Running spam-filter example
Illustrative inputs:
- x1: promotional-language feature
- x2: suspicious-link feature
- x3: untrusted-sender feature
Illustrative weights: w1 = 0.4, w2 = 0.9, w3 = 0.7. Bias: b = -0.5.
The neuron first computes a weighted sum plus bias, then applies an activation function.
Formula
z = Σ(xᵢ × wᵢ) + b
For inputs (1, 1, 1): z = (1×0.4) + (1×0.9) + (1×0.7) - 0.5 = 1.5. Applying sigmoid: σ(1.5) ≈ 0.818.
For inputs (1, 0, 1): z = 0.4 + 0 + 0.7 - 0.5 = 0.6; σ(0.6) ≈ 0.646.
For all-zero inputs: z = -0.5; σ(-0.5) ≈ 0.378.
Important caveats
- Binary feature values are a teaching simplification; real features can be continuous, categorical or otherwise represented numerically.
- “Suspicious link” and “untrusted sender” are illustrative engineered features. The model does not automatically possess those semantic concepts simply because the labels are used.
- Weights and bias are trainable parameters in a typical neuron.
- A neuron's output is a mathematical value passed to another computation or later system component; it is not automatically a business decision.
The exact complete original question sets are not all available in the retained record. The specific final mastery exercise remains outstanding.
Recovered from the available teaching record; not a verbatim transcript.
Introduced; learner mastery not verified.
Help improve this learning path.
Suggest a concept we should teach, or tell us what is unclear. No account is required. Email is optional and only for a reply. Do not include confidential or sensitive information.