In this lesson students are introduced to a form of artificial intelligence called machine learning and how they can use the Problem Solving Process to help train a robot to solve problems. They participate in three machine learning activities where a robot – A.I. Bot – is learning how to detect patterns in fish.

Question of the Day: How can we use the Problem Solving Process to solve a problem with machine learning?

In this lesson students will consider how they create “mental” models when learning new concepts, and how those can be similar to a “machine learning” model. They participate in a color pattern activity to simulate building a machine learning model without help, then they play a game called “Green Glass Door” as an example of supervised learning, and finally, they will sort several scenarios into “supervised” or “unsupervised” learning.

Question of the Day: What are different types of machine learning?

In this lesson, students explore an application of AI called Seeing AI and examine how it is supporting people with visual impairments. Then, students research other examples of how AI is impacting society, focusing on users who are impacted by the examples they find. Finally, students share their findings with each other.

Question of the Day: How is AI helping to solve problems around the world?

In this lesson students will examine several apps that make decisions about what shoes to wear, ultimately building up to an understanding of how machine learning can help make this decision. Students are guided to the conclusion that surveying their users can help them make the best decision by looking for patterns in the data and basing their decisions on these patterns.

Question of the Day: What strategies do computer models use to make decisions?

In this lesson students will participate in an unplugged activity simulating one of the machine learning algorithms computers use to separate data into groups to help make decisions. Students will be tasked with helping a computer learn to classify food as fruits or vegetables, graph 20 different fruits on two axes comparing “sweetness” to “easy to eat”, and then try to separate the data into groups – a fruit area, and a veggie area.

Question of the Day: How do computers learn to classify data?

In this lesson students will dive into the AI Lab tool for the first time, where they select features to train a model that predicts a given label. They start by exploring AI Lab and training a model to recognize shapes. Then they pretend they have been hired by several restaurants who would like to make recommendations to new customers based on survey data they’re collected, go through each dataset, and use data visualization tools to identify features with high relationships in the data.

Question of the Day: How can we use machine learning to make recommendations?

In this lesson students are introduced to importing their models into App Lab and linking their model to their screens. They help create a book recommendation app and learn how to add a welcome screen and events to their code. This lesson assumes students are already familiar with App Lab – for classrooms that have not seen App Lab before, consider extending this lesson and including additional videos or activities that are recommended in the lesson plan.

Question of the Day: How can I create an app using machine learning?

In this lesson, students will investigate a model for bias and be introduced to a Model Card, which is a way of representing important information about a trained model that could help uncover bias. They will be investigating a Medical Priority app, which helps a hospital decide how soon to view patients based on their symptoms. As students go through the activity, they realize that the app is biased based on personal information and examine how this could happen.

Question of the Day: How can we evaluate machine learning models once they’ve been trained?

Students complete the full process of training and saving a model, then importing into App Lab. For the first time, students are able to choose the label they would like to predict and spend time deciding the features they will use to help predict their label of choice. Students also create a model card for their models in order to save them and import it into App Lab

Question of the Day: How can I use Model Cards to document my decisions when training a machine learning model?

In this lesson, students practice importing their models into App Lab, this time including models that have numerical data and using model cards to help improve the user experience of filling out their form. They will then learn how to view the model card within App Lab and use this to add more descriptive elements to an app. Next, they focus on improving the user experience by adding informational text to help guide users through completing the form and adding a style to their app to improve the user experience.

Question of the Day: How can I use a Model Card to improve my app?

In this lesson, students participate in an unplugged activity simulating a zombie outbreak. Students must predict which parts of town have the least amount of zombies using data from a neighboring town. Students will use degrees of similarity and averages to make predictions about the number of zombies at a particular location. Then, students are rescued and get to compare their predictions to the actual numbers as a way to discuss how accuracy is different for numerical data compared to categorical data.

Question of the Day: How do computers learn to make predictions with numerical data?

In this lesson, students will be introduced to numerical data which represents a range of values. Students are presented with a scenario where every feature and label is represented with numerical data, and they learn to use the new data visualization tools within AI Lab to help find patterns.

Question of the Day: How can we use AI Lab to predict numerical data?

In this lesson, students will explore how to customize the code of their app to make additional changes to the design of their app. They will start by exploring a single-screen app and then practice expanding the app to two-screens and updating the code to use the new design mode elements. After this, students help create a Driver Alert app that requires changes to the code using new design mode elements. Using the skills from this lesson, students will be able to create multi-screen apps where questions can appear on multiple screens instead of a single screen.

Question of the Day: How can I customize the code for a machine learning app?

In small groups, students conduct research using articles and videos that expose ethical pitfalls in an Artificial Intelligence (AI) area of their choice. Afterward, each group develops at least one solution-oriented principle that addresses their chosen area. These principles are then assembled into a class-wide “Our AI Code of Ethics” resource (e.g. a slide presentation, document, or webpage) for AI creators and legislators everywhere.

Question of the Day: What are guidelines we can use to create ethical machine learning apps?

In this one or two day project, students apply their skills from the unit so far and create a machine learning app using real-world data. Students are provided with several real-world datasets from a variety of contexts, and they choose which dataset they would like to investigate. They train and save their model, then make a simple App Lab app that uses the model. This mini-project is an opportunity to assess how well students can use features to create accurate machine learning models, and how well they can create apps that use machine learning.

Question of the Day: Can I use real-world data to create an app that uses machine learning?

This is the second in a five-day sequence of lessons that prepare students for the final project. In this lesson, students learn that the other team of students would like to create a club recommender app based on the clubs at their school. Students imagine what questions would be most useful to help make this recommendation, then they learn how to use a Google Form template to create a survey. The steps students take in this lesson are identical to the steps students will take in their final project.

Question of the Day: How can I create a survey to gather data for a machine learning app?

This is the third in a five-day sequence of lessons that prepare students for the final project. In this lesson, students learn how to view survey data in Google Sheets and save the data to their computer as a csv file. Then, they upload the saved data to AI Lab and examine the survey results from one of the students to train a model using their data. Then, students use Google Sheets to examine data from another student where the data has errors and then try to fix the errors. The steps students take in this lesson are identical to the steps students will take in their final project, and the problem-solving strategies they develop will help them overcome challenges in their own final project.

Question of the Day: How can I import data into AI Lab to train a machine learning model?

This is the fourth of a five-day sequence of lessons that prepare students for the final project. In this lesson, students examine survey data from other members of the student team and analyze why their models are not working correctly. In examining the data, students develop strategies for avoiding these issues in the future and strategies for coping with these issues should they happen again. These are skills students will use in the final project as they develop their own surveys and collect data.

Question of the Day: What are strategies to make sure our data generates an accurate model?

This is the fifth of a five-day sequence of lessons that prepare students for the final project. In this lesson, students import the club recommender app into App Lab and begin customizing the app. Students add a welcome screen and update the descriptions of each feature, then they can decide how they would like to further customize the app. The steps students take in this lesson are identical to the steps students will take in their final project.

Question of the Day: How can I create a friendly, easy-to-use machine learning app?

To conclude this unit, students develop an AI app that addresses the social issue of their interest. Students follow a project guide to complete this multi-day activity. In the first step, students prepare the data they will use to train their model in AI Lab. After training, testing, and generating a model card, they export their model into App Lab for development. Here they use their model to create a user-friendly app. Students perform a peer review and make any necessary updates to their projects while reflecting on the outcome.

Question of the Day: How can I create an AI App the solves a problem in my community?