Neuron

Definition

A neuron is a small computation unit inside a neural network.

It takes input values, combines them with parameters, and produces an output.

Basic pattern:

inputs -> weighted combination -> activation -> output

A neuron is a parameterized input-output computation.

---

Why It Exists

A single neuron gives the network a reusable unit of transformation.

It can ask:

which inputs matter?
how strongly should they matter?
should the result be passed forward?

The neuron does this using weights, bias, and an activation function.

A neuron turns several input values into one transformed value.

---

Simple Example

Suppose a neuron receives two inputs.

input 1 = 3
input 2 = 5

It may assign different importance to each input.

input 1 matters strongly
input 2 matters weakly

The neuron combines them into a single score.

combined score -> output

A neuron compresses multiple signals into one computed signal.

---

Neuron vs Biological Neuron

The word "neuron" is inspired by biology, but in this context it is a mathematical unit.

It is not a full biological model.

Here, a neuron means:

an input-output function with adjustable parameters

In neural networks, a neuron is a computation, not a biological explanation.

---

What Is Adjustable

A neuron has parameters.

Usually these include:

weights
bias

The inputs change from example to example.

The parameters are adjusted during training.

Training changes the neuron by changing its parameters.

---

Final Summary

• A neuron is a small computation unit in a neural network.
• It receives inputs and produces one output.
• It uses weights, bias, and an activation function.
• It is a mathematical unit, not a full biological model.
• Inputs are processed; parameters are adjusted.
• Networks are built by connecting many such units.