The proposed machine learning lesson plan focuses on the competency of "AI techniques and applications". It begins by introducing students to how machine learning works and how it contributes to the development of artificial intelligence. Students then engage in a hands-on activity where they train their own AI to recognize images or audio. As an extension, teachers can guide students in designing a project where they teach their AI something they believe represents a useful or creative application of machine learning.
You can download the lesson plan pdf here.
Grade Level: 10-12
Time Required: 45-60 minutes
By the end of this lesson, students will be able to:
Explain the basics of machine learning and how models are trained.
Identify how bias in data affects machine learning outcomes.
Discuss the impact of machine learning on society, particularly in environmental conservation.
Apply machine learning concepts by training a simple model using Google's Teachable Machine.
Computers with internet access
Videos from code.org:
What is Machine Learning?
Training Data and Bias
Impact on Society
Printed images of various ocean animals and garbage (There are comprehensive data sets of images already compiled on www.kaggle.com )
Access to Teachable Machine
Activity: Watch the first video on What is Machine Learning.
Discussion Questions:
How does machine learning differ from traditional programming?
What kinds of tasks can machine learning accomplish?
Where have you encountered machine learning in your daily life?
Step 1: Set Up the Teachable Machine Model
Guide students to open Teachable Machine and create a new image classification project.
Instruct students to create two classes: "Fish" and "Garbage".
Upload images of different fish species under "Fish" and various trash items under "Garbage".
Train the model and test it with new images.
Step 2: Introducing Data Bias
Show students a printed image of an octopus and ask them to predict what the AI will classify it as.
Observe and discuss the results.
Discussion Questions:
Did the AI classify the octopus incorrectly? Why?
What does this tell us about bias in training data?
How can we improve the accuracy of our model?
Step 3: Expanding the Training Data
Add new categories: "Octopus", "Dolphin", "Crab", etc.
Train the model again and test with more images.
Observe improvements and discuss why having diverse data makes AI smarter.
Activity: Watch the second and third videos on Training Data Bias and Impact on Society.
Discussion Questions:
How does bias in data affect AI decision-making in real-world applications?
How could this bias impact different communities?
How can AI be used to help environmental conservation, such as cleaning up the ocean?
What are the ethical responsibilities of AI developers?
Exit Ticket: Students write a short response to:
"If you were designing an AI system to help clean the ocean, what would you need to consider when training the model?"
Have students create and train their own AI for a specific purpose
The projects already on the "Teachable Machine" website might help inspire some ideas