Lesson plan introduction
Welcome to the " Machine Learning with Scratch" lesson plan. In six (6) courses, we will introduce primary school students to hands-on experience of Artificial Intelligence's magical and playful world.
During this module, students can learn the basics of machine learning and coding by creating projects and building games with the machine learning models they've trained.
Lessons in this course are based on the gamification method. Students will use the game development app, “Scratch”, to create games while learning to use machine learning and coding. Students will first teach the machine learning engine by inserting data inputs via sound, text, images or webcam images and exporting the outputs of the project in the “Scratch” platform to incorporate into their code.
Machine Learning with Scratch
1. Smart Pet
In this lesson, students will develop a project to train their virtual Scratch pet to identify and perform voice commands! The educator will guide the students on how to input sound data into the machine-learning machine and export the outputs on the “Scratch” platform. After the training, students can use the machine learning command blocks and train their virtual pet to react to their voice commands, learning basic programming concepts like sequencing, conditions, events, and responses simultaneously.
2. The Maze
The Maze is an interactive lesson in which students create a maze game and use voice commands to guide their virtual character to follow the correct path in order to escape. The course will start with the students applying the knowledge from the first lesson about the correct way to input sound data into the machine learning machine and train the engine. Then, students will program their virtual characters to perform the voice commands to move throughout the maze and escape.
3. Recycling Game
Teaching recycling lessons in primary school is a valuable way to instil environmental awareness and responsibility in young students. With this lesson, students will create a valuable and enjoyable game that classifies recyclable and non-recyclable items by presenting them to the camera. The teacher guides the students to train the machine learning engine via webcam inputs of the everyday objects students use at school. Then, we use the outputs and program our character to throw the items we present to the webcam into the correct bin.
4. Up & Down Game
Students are challenged to create an even more interactive game with moving objects except for the main character. Students must program moving objects that interact with the main character following the “object-oriented programming” method. The virtual character will be programmed to dodge the obstacles and will be controlled with voice commands inputted from the machine learning engine. The students will have to complete the game by putting a counting variable for score and high score, learning to use variables simultaneously. The final challenge of this lesson is the insertion of score and high-score variables into the game. To finalize and complete the project, students will be introduced to the variable concept and will use them to count their scores.
5. Smart Home Assistant
The smart home is a usual and modern application of AI nowadays. With this project, students can develop and program a virtual smart home that their virtual assistant will control. The virtual assistant will be trained with the machine learning machine to follow text or voice commands and operate the house objects.
6. Virtual Calculator
Students will finalize the course by developing a fully functional adder calculator. To create this calculating tool, the teacher guides students to train the machine learning engine to identify the sum of different attends from 0 to 10 via text. At the end of the project, the calculator should be able to identify the numbers of the text and present the sum in the background. This lesson will complete the learning of the usage of AI and test the knowledge students obtained during the whole course of AI.