🎯 What you will learn
- Rules are written by people
- Learning finds patterns in examples
- Most modern AI learns from data
👋 Let’s get started!
Traditional software follows rules that a person wrote. Machine learning is different: it looks at many examples and finds the pattern itself. Let’s compare the two. 🧠
🪜 Step by step
1. 📜 Hand-written rules
A person decides every step. This always works the same way.
def is_fruit(name):
return name in ["apple", "banana"]
print(is_fruit("apple"))2. 📚 Learned from examples
The program gets examples with answers and works out its own pattern.
examples = [(1, "cat"), (2, "cat"), (8, "dog"), (9, "dog")]3. 🔍 Patterns, not facts
A learned pattern is only as good as its examples.
print(len(examples), "examples")🧪 Try it!
Change the code and press ▶ Run to see what happens. You can’t break anything!
Rules vs examples
def is_fruit(name):
return name in ["apple", "banana"]
print(is_fruit("apple"))
print(is_fruit("car"))
examples = [(1, "cat"), (2, "cat"), (8, "dog"), (9, "dog")]
print(len(examples), "examples")🎨 Unplugged activity
🧠 Write down 5 things you think a phone camera “knows” about your face. For each one, ask: was it a rule someone wrote, or a pattern from examples?
Rule or learning?
# Rule: if the number is over 10, say big
def size(n):
return "big" if n > 10 else "small"
print(size(12))📖 Key words
Tap a card to flip it over!
- Rule
- A step a person writes 📜
- Machine learning
- Finding patterns from examples 🔍
- Example
- A piece of data with an answer 📚
🏆 Challenge
Print how many examples are in the list “examples” (three of them). Press ▶ Run.
Challenge
examples = [(1, "cat"), (2, "cat"), (8, "dog")]
Knowledge Check Quiz
- What does machine learning use to find patterns?
- Which is hand-written?
- Why can a learned pattern be wrong?