← Back to Home Tech & AI Classes

How Do Machines Learn? 🤖

Understanding the basics of machine learning
Teacher: Mr. Yousef Younis
"AI doesn't replace human intelligence — it enhances it."

Types of Machine Learning 🧠

📚 Supervised Learning

Uses labeled data (inputs with correct answers)

  • AI learns by example
  • Like identifying cats vs. dogs
  • Used in spam detection & diagnosis
🔍 Unsupervised Learning

Finds patterns in unlabeled data

  • Groups similar items together
  • No teacher needed
  • Used in music & photo apps
🎮 Reinforcement Learning

Learns through trial and error

  • Uses rewards & punishments
  • Like teaching a robot to walk
  • Used in gaming & robotics

Inside a Neural Network 🧩

Network Structure

  • Input Layer: Receives raw data
  • Hidden Layers: Process information
  • Output Layer: Produces results

Key Ideas

  • Each neuron does a small task
  • Networks learn patterns over time
  • Inspired by human brain structure

Real-World Example

Handwriting Recognition:

  • Input: Scanned digit image
  • Hidden Layers: Analyze shapes
  • Output: Predicted number (0-9)

Bias in Artificial Intelligence ⚖️

What Is Bias?

Unfair favoritism in AI systems, often unintentional but impactful

Common Types of Bias

  • Language bias
  • Racial bias
  • Gender bias
  • Age bias
  • Economic bias

Solutions

  • Use diverse datasets
  • Include different perspectives
  • Regular fairness testing
  • Continuous improvement

Case Study: Joy Buolamwini

Discovered facial recognition systems worked better for lighter-skinned faces, leading to important discussions about AI fairness.

Final Review 📘

When Analyzing an AI System:

  • Identify the learning type (Supervised/Unsupervised/Reinforcement)
  • Understand neural network involvement
  • Check for potential bias issues
  • Consider ethical implications
"The goal isn't just to build smart AI, but to build fair and ethical AI that benefits everyone."
↓ Scroll or use arrow buttons