DAY 1 · RULES & PATTERNS

Monster Club

Who gets in tonight?
Follow a rule. Then predict a machine.

Two rounds · About 9 to 12 minutes

FROM THE DAY 1 LESSON

Two ways to make a decision

You’ll try both at the club door.

A person writes a rule on the left. A machine uses a pile of examples on the right.
A person writes the rules.
A machine learns from examples.
ROUND 1 · FOLLOW A RULE

You work the door

Read the owner’s rule. Choose IN or OUT for each monster before time runs out.

ROUND 2 · FIND A PATTERN

Predict the machine

Study labeled photos. Predict whether the machine will let a new monster IN or send it OUT.

Keyboard: ← IN · → OUT

ROUND 1 · RULE-BASED

Read the owner’s rule

A PERSON WROTE THIS
Has horns? Let it IN.
No horns? Send it OUT.
InputA monster arrives.
RuleCheck for horns.
OutputIN or OUT.

You’re acting as a rule-based machine: follow the instructions exactly.

5 seconds per monster. Fast, correct choices earn more points.

ROUND 1 COMPLETE

You followed a written rule

The helmet exposed a missing detail.
A person fixed it by changing the instructions.

ROUND 2 · MACHINE LEARNING

Now the machine decides

In this simulation, a door machine uses patterns from 12 example photos. Each has a label: IN or OUT.

1. StudyLook at the labeled photos.
2. PredictChoose what the machine will say.
3. WatchSee whether you matched it.

Just 4 predictions, then see how the model decides. No timer.

TRAINING DATA = LABELED EXAMPLES

Who got in last night?

Look for patterns in the horns, spots, eyes, and hats.

20 seconds to study
FROM EXAMPLES TO A MODEL

Learning from examples…

Round 1Score 0Streak 0
THE RULE CARDIf a monster has horns, let it IN.
THE MACHINE SAYS

ROUND 2 COMPLETE

Does one of these rules explain it?

Pick a rule. Watch us compare it with all 12 labels.

INSIDE THIS MODEL

Several features work together

Those three rules miss some examples. This model combines four features.

Add the points: 1 or more → IN.Less than 1 → OUT.

In machine learning, training adjusts which features matter and how much.
This game uses preset points to show how a simple model can combine features.

YOUR NIGHT AT MONSTER CLUB

Closing time!

Follow a written rule

A person wrote the instructions.
You fixed the missing detail.

Predict a model

You studied examples and predicted how the machine would classify new monsters.

Data

The examples used for learning.

Model

What the machine uses after learning.

Classify

Put something into a group.