DEEP LEARNING LESSON 1 FOUNDATIONS

What Is a Neural Network?

A neural network is a Machine Learning model inspired by the way connected neurons work together. It uses layers of connected nodes to learn patterns from data and make predictions.

The simple idea

A neural network takes input data, processes it through connected layers, and produces an output. During training, it adjusts its internal values so that its predictions become more accurate.

What Is a Neural Network?

A neural network is a model made up of many small processing units called neurons.

These neurons are organized into layers. Each layer receives information, processes it, and passes the result to the next layer.

INPUT
Hours Studied
Attendance
Previous Score
HIDDEN LAYERS
Learn Patterns
Combine Information
OUTPUT
Pass / Fail

The network learns how the input information is related to the desired output.

Basic Structure of a Neural Network

A basic neural network contains three important parts: an input layer, one or more hidden layers, and an output layer.

Input Layer
Hidden Layer(s)
Output Layer

The input layer receives the data. Hidden layers process the information and learn patterns. The output layer produces the final prediction.

Example: Predicting Whether a Student Will Pass

Suppose we want to predict whether a student will pass an exam.

We provide three pieces of information:

Hours Studied
Attendance
Previous Score

These values become the inputs to the neural network.

Student Data
Neural Network
Pass

During training, the neural network looks at many examples of students and learns relationships between their information and their results.

What Is a Neuron?

A neuron is a small processing unit inside a neural network. It receives values, applies weights and a bias, performs a calculation, and produces an output.

Input 1
+
Input 2
+
Input 3
Neuron
Output

The neuron does not simply treat every input as equally important. The weights determine how strongly each input affects the calculation.

Why Do Neurons Have Weights?

Different inputs can have different levels of importance.

For example, when predicting whether a student will pass, previous exam performance might be more useful than another input.

Simple Example

Hours Studied Weight
Attendance Weight
Previous Score Weight

During training, the network adjusts these weights so that its predictions improve.

How Does Information Move Through the Network?

Information moves from one layer to the next. This process is called forward propagation.

Input
Input Layer
Hidden Layer
Hidden Layer
Output

Each layer performs calculations and passes its results to the next layer.

How Does a Neural Network Learn?

A neural network does not know the correct weights when it starts training.

It starts with initial values, makes predictions, compares those predictions with the correct answers, and then adjusts its weights.

1. Input Give data to the network.
2. Prediction Network produces an answer.
3. Error Compare prediction with the correct answer.
4. Update Adjust the weights.

This process is repeated many times until the model learns useful patterns from the training data.

Example: Recognizing a Cat

A neural network can also process images. An image contains many pixel values that become the input.

Image Pixels
Early Patterns
Shapes
Object Features
Cat

The network can learn increasingly complex representations through its layers.

Why Is It Called a Neural Network?

The word "network" comes from the way many neurons are connected together.

INPUT
Input A
Input B
NEURONS
Neuron 1
Neuron 2
Neuron 3
OUTPUT
Prediction

The connections allow information to flow through the network and allow the model to learn relationships between inputs and outputs.

Important

Neural networks are inspired by biological neurons, but they are not actual human brains. They are mathematical models implemented using computers.

QUICK CHECK

Check Your Understanding

What is a neural network?
A Machine Learning model made of connected processing units called neurons.

What are the main parts?
An input layer, hidden layers, and an output layer.

What does a neuron use to calculate its output?
Inputs, weights, a bias, and a mathematical activation process.

How does the network learn?
It makes predictions, measures errors, and adjusts its weights during training.

NEXT TOPIC

How Deep Learning Learns

Now we will look more closely at how a Deep Learning model learns patterns from training data.