Decision Tree Workshop

Build your own AI classifier — using nothing but questions!

🔍 What Is a Decision Tree? A decision tree is a series of YES/NO questions that an AI uses to sort things into groups. Each question sends you down a different path — until you reach an answer. Kai the robot built one to classify animals. Now it's your turn to build one from scratch, using your own questions. Any questions you choose!

📦 Your Training Animals These are the 8 animals your tree must correctly classify. Study them carefully before choosing your questions.

🐶 Dog

🐱 Cat

🐟 Fish

🦅 Eagle

🐴 Horse

🐋 Whale

🦊 Fox

🐦 Bird

🎯 Your Mission Step 1: Choose 3 YES/NO questions that split these 8 animals into 4 groups.
Step 2: Draw your decision tree on Page 2 — one question per level.
Step 3: On Page 3, test your tree on 5 animals your tree has never seen. Does it work?

🧠 Choosing Good Questions A good decision tree question is a yes/no question that splits the animals into roughly equal groups. "Does it live in water?" splits 8 animals into 2 water and 6 land — a bit uneven. "Does it have fur?" might split them more evenly. Try a few questions on paper before committing. The goal: every animal ends up in a different final group, with no confusion!

🤖 Kai says: "I learned to classify animals by studying examples — exactly what you're doing now. The difference is I needed programmers to design my questions first. You're the programmer today!"

HiKIDAI Activity 01 · My Decision Tree

📋 Build your decision tree below. Write your question in each box, then write the animal group names in the leaves.

Tip: Check each of your 8 training animals — make sure every single one lands in exactly one leaf.

🔍 Root Question (Question 1):

YES ✅

Question 2 (if YES above):

YES

Group Name:

Animals:

NO

Group Name:

Animals:

NO ❌

Question 3 (if NO above):

YES

Group Name:

Animals:

NO

Group Name:

Animals:

✅ Self-Check Before Moving On Run all 8 training animals through your tree. Does each one end in exactly one leaf? If two animals end up in the same leaf with different labels, your questions need adjusting. That's normal — real AI engineers test and revise too!

HiKIDAI Activity 01 · Test Your Decision Tree!

🧪 Phase 3 — Test Your Tree

Test Cases — Animals Your Tree Has Never Seen!

Run each of these animals through your decision tree on Page 2. Answer all 3 questions in order, follow the YES/NO paths, and write which group the animal lands in. Then say whether you think your tree got it right!

🐧 Penguin

🦇 Bat

🦁 Lion

🦈 Shark

🦜 Parrot

Animal	Q1 Answer (Y/N)	Q2 or Q3 Answer	Tree puts it in…	Is the tree right?
🐧 Penguin				⬜ Yes ⬜ No
🦇 Bat				⬜ Yes ⬜ No
🦁 Lion				⬜ Yes ⬜ No
🦈 Shark				⬜ Yes ⬜ No
🦜 Parrot				⬜ Yes ⬜ No

🤔 Reflection Which animal confused your tree the most? Why do you think your questions didn't work for it?

💡 The Big Insight If your tree fails on Penguin (can't fly, no fur, not a water-only animal), that's because Penguin wasn't in your training data. Incomplete training data = confused AI.