raspberrypilearning · sashamishcheriakova · Feb 25, 2025 · Feb 25, 2025 · Feb 25, 2025 · Feb 25, 2025
diff --git a/ar-SA/images/banner.png b/ar-SA/images/banner.png
diff --git a/ar-SA/images/decision1.png b/ar-SA/images/decision1.png
diff --git a/ar-SA/images/decision2.png b/ar-SA/images/decision2.png
diff --git a/ar-SA/images/final-tree.png b/ar-SA/images/final-tree.png
diff --git a/ar-SA/images/whatyouwillmake.png b/ar-SA/images/whatyouwillmake.png
diff --git a/ar-SA/meta.yml b/ar-SA/meta.yml
@@ -0,0 +1,27 @@
+title: Dinosaur decision tree
+hero_image: images/banner.png
+description: Create a paper-based decision tree to classify dinosaurs, and learn how machine learning models use decision trees to classify data
+version: 1
+listed: true
+copyedit: false
+last_tested: "2024-06-27"
+steps:
+  - 
+    title: What you will make
+  - 
+    title: Classify the dinosaurs
+    completion:
+      - engaged
+  - 
+    title: Modify the criteria
+  - 
+    title: More than one question
+    completion:
+      - internal
+  - 
+    title: Decision tree
+    challenge: true
+  - 
+    title: What can you do now?
+    completion:
+      - external
diff --git a/ar-SA/resources/NEW Dino 1.srt b/ar-SA/resources/NEW Dino 1.srt
@@ -0,0 +1,32 @@
+1
+00:00:04,433 --> 00:00:07,866
+You can use a decision tree to categorise dinosaurs.
+
+2
+00:00:08,032 --> 00:00:10,165
+This dinosaur is a Theropod
+
+3
+00:00:10,732 --> 00:00:12,831
+and this one is a Sauropod.
+
+4
+00:00:12,998 --> 00:00:16,097
+To categorise these two dinosaurs you could ask the question:
+
+5
+00:00:16,164 --> 00:00:18,399
+"Was it longer than 6 metres?"
+
+6
+00:00:20,396 --> 00:00:24,762
+Given only these two dinosaurs, if it's longer than 6 metres it's a Sauropod
+
+7
+00:00:24,762 --> 00:00:27,195
+and if it's not, it's a Theropod.
+
+8
+00:00:30,228 --> 00:00:34,860
+Think of a different question you could ask to separate out these two categories of dinosaur.
+
diff --git a/ar-SA/resources/NEW Dino 2.srt b/ar-SA/resources/NEW Dino 2.srt
@@ -0,0 +1,18 @@
+1
+00:00:03,800 --> 00:00:07,320
+But what if we add another dinosaur?
+
+2
+00:00:07,320 --> 00:00:12,960
+Allosaurus is longer than 
+6 metres, so it goes here:
+
+3
+00:00:12,960 --> 00:00:18,760
+but it's a Theropod, so it should go here.
+
+4
+00:00:18,760 --> 00:00:27,600
+What part of the question would you change 
+to make the decision tree work again?
+
diff --git a/ar-SA/resources/NEW Dino 3.srt b/ar-SA/resources/NEW Dino 3.srt
@@ -0,0 +1,28 @@
+1
+00:00:04,480 --> 00:00:06,880
+What happens if we add another dinosaur?
+
+2
+00:00:06,880 --> 00:00:15,000
+The question "Did it live more than 130 million 
+years ago?" splits the dinosaurs like this:
+
+3
+00:00:15,000 --> 00:00:16,560
+That's not much use!
+
+4
+00:00:16,560 --> 00:00:20,880
+But even if we switch the question 
+to '152 million years ago',  
+
+5
+00:00:20,880 --> 00:00:29,600
+we still can't categorise the 
+dinosaurs using only one question.
+
+6
+00:00:29,600 --> 00:00:42,920
+What question could you use to 
+categorise the remaining dinosaurs?
+
diff --git a/ar-SA/resources/NEW Dino 4.srt b/ar-SA/resources/NEW Dino 4.srt
@@ -0,0 +1,22 @@
+1
+00:00:04,360 --> 00:00:09,600
+A machine learning model that uses a decision 
+tree must repeatedly refine its criteria.
+
+2
+00:00:09,600 --> 00:00:12,880
+The more data input, the more accurate it becomes.
+
+3
+00:00:12,880 --> 00:00:15,040
+This is called training.
+
+4
+00:00:15,040 --> 00:00:18,160
+Use what you have learned to draw 
+a decision tree that allows you  
+
+5
+00:00:18,160 --> 00:00:29,720
+to correctly identify each category of dinosaur.
+
diff --git a/ar-SA/resources/readme.txt b/ar-SA/resources/readme.txt
@@ -0,0 +1 @@
+To watch a video with subtitles on VLC (videolan.org), ensure the video file and subtitle file are in the same folder and have the exact same name (e.g., video.mp4 and video.srt). Open the video in VLC, and it will automatically load the subtitles. If the subtitles don’t appear, right-click the video screen, go to **Subtitle**, then **Add Subtitle File**, and select the correct .srt file. Enjoy watching with subtitles!
diff --git a/ar-SA/step_1.md b/ar-SA/step_1.md
@@ -0,0 +1,22 @@
+## What you will make
+
+Some machine learning models use **decision trees** to classify data.
+
+Create your own paper-based decision tree to classify dinosaurs, and learn about how a model is trained.
+
+![Eleven 'cards' with dinosaur pictures and facts about the dinosaurs](images/whatyouwillmake.png)
+
+## --- collapse ---
+
+## title: No YouTube? Download the video!
+
+You can download all the videos for this project, including the above video [by clicking here](https://rpf.io/p/en/decision-tree-go){:target="_blank"}.
+
+\--- /collapse ---
+
+### Data
+
+The data in this project is from the Natural History Museum's [Dino Directory](https://www.nhm.ac.uk/discover/dino-directory.html){:target="_blank"}.
+
+Dinosaur illustrations:
+Allosaurus, Brachiosaurus - Pixabay. Archaeoceratops - Nobu Tamura (CC BY 3.0). Bambiraptor - Conty (CC BY-SA 4.0). Chindesaurus - Jeff Martz (Public domain). Concavenator, Maiasaura - UnexpectedDinoLesson (CC BY-SA 4.0). Diplodocus, Herrerasaurus - Fred Wierum (CC BY-SA 4.0). Parksosaurus - Steveoc 86 (CC BY-SA 3.0). Zephyrosaurus - Emily Willoughby (CC BY-SA 3.0).
diff --git a/ar-SA/step_2.md b/ar-SA/step_2.md
@@ -0,0 +1,43 @@
+## Classify the dinosaurs
+
+<html>
+  <div style="position: relative; overflow: hidden; padding-top: 56.25%;">
+    <iframe style="position: absolute; top: 0; left: 0; right: 0; width: 100%; height: 100%; border: none;" src="https://www.youtube.com/embed/3op4RCy1wRc?rel=0&cc_load_policy=1" allowfullscreen allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share"></iframe>
+  </div>
+</html>
+
+<p style="border-left: solid; border-width:10px; border-color: #0faeb0; background-color: aliceblue; padding: 10px;">
+**Aim of the project:** Each dinosaur has a <span style="color: #0faeb0">**category**</span>. A category describes a group of dinosaurs with similar characteristics. You need to tell which category a dinosaur is in, using the facts you have about the dinosaur.
+</p>
+
+Here are some facts about two different dinosaurs:
+
+(mya = million years ago)
+
+| Name         | Length (m) | Diet        | Continent     | Lived (mya) | Category |
+| ------------ | ----------------------------- | ----------- | ------------- | ------------------------------ | -------- |
+| Concavenator | 6                             | Carnivorous | Europe        | 130                            | Theropod |
+| Diplodocus   | 26                            | Herbivorous | North America | 152                            | Sauropod |
+
+You could separate this data into the two <span style="color: #0faeb0">**categories**</span> of dinosaur by asking this question:
+
+![Image of a decision tree with the question 'Was it longer than 6 metres?'](images/decision1.png)
+
+If the answer is **yes**, the dinosaur must be a Sauropod, and if it is **no** then it must be a Theropod.
+
+\--- task ---
+
+Think of a different question you could ask to separate out these two categories of dinosaur.
+
+## --- collapse ---
+
+## title: Show me the answer
+
+- Did it live in North America?
+- Was it carnivorous?
+- Does its name begin with 'C'?
+- Did it live more than 130 million years ago?
+
+\--- /collapse ---
+
+\--- /task ---
diff --git a/ar-SA/step_3.md b/ar-SA/step_3.md
@@ -0,0 +1,31 @@
+## Modify the criteria
+
+<html>
+  <div style="position: relative; overflow: hidden; padding-top: 56.25%;">
+    <iframe style="position: absolute; top: 0; left: 0; right: 0; width: 100%; height: 100%; border: none;" src="https://www.youtube.com/embed/zYh4q8dVLw4?rel=0&cc_load_policy=1" allowfullscreen allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share"></iframe>
+  </div>
+</html>
+
+Let's add the data for another dinosaur.
+
+| Name         | Length (m) | Diet        | Continent     | Lived (mya) | Category |
+| ------------ | ----------------------------- | ----------- | ------------- | ------------------------------ | -------- |
+| Allosaurus   | 12                            | Carnivorous | Europe        | 152                            | Theropod |
+| Concavenator | 6                             | Carnivorous | Europe        | 130                            | Theropod |
+| Diplodocus   | 26                            | Herbivorous | North America | 152                            | Sauropod |
+
+The question **"Was it longer than 6 metres?"** no longer helps you to separate the dinosaurs into the two categories, because the Allosaurus is a Theropod, but it was 12m long.
+
+\--- task ---
+
+What part of the question would you change to make it work again?
+
+## --- collapse ---
+
+## title: Show me the answer
+
+Change the size and ask "Was it longer than **12** metres?"
+
+\--- /collapse ---
+
+\--- /task ---
diff --git a/ar-SA/step_4.md b/ar-SA/step_4.md
@@ -0,0 +1,56 @@
+## More than one question
+
+<html>
+  <div style="position: relative; overflow: hidden; padding-top: 56.25%;">
+    <iframe style="position: absolute; top: 0; left: 0; right: 0; width: 100%; height: 100%; border: none;" src="https://www.youtube.com/embed/1YGBq7GWiZU?rel=0&cc_load_policy=1" allowfullscreen allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share"></iframe>
+  </div>
+</html>
+
+Let's add another dinosaur:
+
+| Name          | Length (m) | Diet        | Continent     | Lived (mya) | Category       |
+| ------------- | ----------------------------- | ----------- | ------------- | ------------------------------ | -------------- |
+| Allosaurus    | 12                            | Carnivorous | Europe        | 152                            | Theropod       |
+| Concavenator  | 6                             | Carnivorous | Europe        | 130                            | Theropod       |
+| Diplodocus    | 26                            | Herbivorous | North America | 152                            | Sauropod       |
+| Herrerasaurus | 3                             | Carnivorous | South America | 228                            | Early dinosaur |
+
+\--- task ---
+
+Is it possible for you to split the dinosaurs into categories using the question **"Did it live more than 130 million years ago?"**?
+
+## --- collapse ---
+
+## title: Show me the answer
+
+No, because there are now three dinosaurs that lived more than 130 million years ago, but they are in different categories.
+
+Even if you changed the criteria to **more than 152 million years ago**, you still can't separate them into three categories.
+
+\--- /collapse ---
+
+\--- /task ---
+
+To put dinosaurs in three categories, you need to add another question.
+
+\--- task ---
+
+Choose a **different** piece of data from the table; for example, length or diet.
+
+What question could you write using this data in the blank space to correctly categorise the remaining dinosaurs?
+
+![A decision tree with the first question 'Did it live more than 152 million years ago?' with yes = Early dinosaur and No = a blank space for a second question.](images/decision2.png)
+
+## --- collapse ---
+
+## title: Show me the answer
+
+Any of the following questions would give the correct category for every dinosaur:
+
+- Was it shorter than 26 metres?
+- Was it a carnivore?
+- Did it live in Europe?
+
+\--- /collapse ---
+
+\--- /task ---
diff --git a/ar-SA/step_5.md b/ar-SA/step_5.md
@@ -0,0 +1,59 @@
+## Decision tree
+
+\--- challenge ---
+
+<html>
+  <div style="position: relative; overflow: hidden; padding-top: 56.25%;">
+    <iframe style="position: absolute; top: 0; left: 0; right: 0; width: 100%; height: 100%; border: none;" src="https://www.youtube.com/embed/mYkxL-efCIs?rel=0&cc_load_policy=1" allowfullscreen allow="accelerometer; autoplay; clipboard-write; encrypted-media; gyroscope; picture-in-picture; web-share"></iframe>
+  </div>
+</html>
+
+A machine learning model that uses a decision tree must repeatedly refine its criteria. The more data input, the more accurate it becomes — this is called **training**. You have only used a few criteria, but a machine learning model might use many **thousands** of values.
+
+Here is a larger set of data about dinosaurs:
+
+(mya = million years ago)
+
+| Name            | Length (m) | Diet        | Continent     | Lived (mya) | Category       |
+| --------------- | ----------------------------- | ----------- | ------------- | ------------------------------ | -------------- |
+| Allosaurus      | 12                            | Carnivorous | Europe        | 152                            | Theropod       |
+| Archaeoceratops | 1.3           | Herbivorous | Asia          | 121                            | Ceratopsian    |
+| Bambiraptor     | 1                             | Carnivorous | North America | 84                             | Theropod       |
+| Brachiosaurus   | 30                            | Herbivorous | North America | 155                            | Sauropod       |
+| Chindesaurus    | 4                             | Carnivorous | North America | 227                            | Early dinosaur |
+| Concavenator    | 6                             | Carnivorous | Europe        | 130                            | Theropod       |
+| Diplodocus      | 26                            | Herbivorous | North America | 152                            | Sauropod       |
+| Herrerasaurus   | 3                             | Carnivorous | South America | 228                            | Early dinosaur |
+| Maiasaura       | 9                             | Herbivorous | North America | 80                             | Ornithopod     |
+| Parksosaurus    | 3                             | Herbivorous | North America | 76                             | Ornithopod     |
+| Zephyrosaurus   | 1.8           | Herbivorous | North America | 120                            | Ornithopod     |
+
+<p style="border-left: solid; border-width:10px; border-color: #0faeb0; background-color: aliceblue; padding: 10px;">
+Why not download and print these [dinosaur cards](resources/dinosaur_cards.pdf){:target="_blank"} and use them to help you draw the decision tree?
+</p>
+
+\--- task ---
+
+Draw a decision tree that allows you to correctly identify each <span style="color: #0faeb0">**category**</span> of dinosaur.
+
+**Tip:** Each question should split the data so that one category of dinosaur is fully identified.
+
+\--- /task ---
+
+\--- task ---
+
+[Choose another dinosaur](https://www.nhm.ac.uk/discover/dino-directory.html){:target="_blank"} and use your decision tree to identify which category it is in. Was your decision tree correct?
+
+\--- /task ---
+
+## --- collapse ---
+
+## title: Show me the answer
+
+Here is one possible solution, but there are many valid trees you could draw:
+
+![Final decision tree with questions: Was it less than 26m long? Did it live less than 227m years ago? Was it carnivorous?](images/final-tree.png)
+
+\--- /collapse ---
+
+\--- /challenge ---
diff --git a/ar-SA/step_6.md b/ar-SA/step_6.md
@@ -0,0 +1,3 @@
+## What can you do now?
+
+There are lots of other machine learning and AI projects in the [artificial intelligence](https://projects.raspberrypi.org/en/pathways/ai-toolkit) toolkit.
diff --git a/cy-GB/images/banner.png b/cy-GB/images/banner.png
diff --git a/cy-GB/images/decision1.png b/cy-GB/images/decision1.png
diff --git a/cy-GB/images/decision2.png b/cy-GB/images/decision2.png
diff --git a/cy-GB/images/final-tree.png b/cy-GB/images/final-tree.png
diff --git a/cy-GB/images/whatyouwillmake.png b/cy-GB/images/whatyouwillmake.png
diff --git a/cy-GB/meta.yml b/cy-GB/meta.yml
@@ -0,0 +1,27 @@
+title: Dinosaur decision tree
+hero_image: images/banner.png
+description: Create a paper-based decision tree to classify dinosaurs, and learn how machine learning models use decision trees to classify data
+version: 1
+listed: true
+copyedit: false
+last_tested: "2024-06-27"
+steps:
+  - 
+    title: What you will make
+  - 
+    title: Classify the dinosaurs
+    completion:
+      - engaged
+  - 
+    title: Modify the criteria
+  - 
+    title: More than one question
+    completion:
+      - internal
+  - 
+    title: Decision tree
+    challenge: true
+  - 
+    title: What can you do now?
+    completion:
+      - external
Original file line number	Diff line number	Diff line change
		@@ -0,0 +1 @@
		To watch a video with subtitles on VLC (videolan.org), ensure the video file and subtitle file are in the same folder and have the exact same name (e.g., video.mp4 and video.srt). Open the video in VLC, and it will automatically load the subtitles. If the subtitles don’t appear, right-click the video screen, go to Subtitle, then Add Subtitle File, and select the correct .srt file. Enjoy watching with subtitles!
Original file line number	Diff line number	Diff line change
		@@ -0,0 +1,3 @@
		## What can you do now?

		There are lots of other machine learning and AI projects in the [artificial intelligence](https://projects.raspberrypi.org/en/pathways/ai-toolkit) toolkit.