AI & Machine Learning · Course Guide
Hands-On AI & Machine Learning for Kids in Stittsville, Ottawa & Online: Train Smart Vision & Audio Models
Demystifying artificial intelligence: how kids train vision and audio neural models, discover dataset bias, and become thoughtful AI creators rather than mere consumers.
📍 In-Person Classes in Stittsville, Ottawa & 🌐 Live Online Anywhere
Whether your family prefers an intimate physical classroom setting in Stittsville, Ottawa or the comfort of a live interactive online classroom from home anywhere, ReaderNook offers ultra-small class sizes with dedicated instructor mentorship.
- In-Person Cohorts (Stittsville, Ottawa): Strict cap of 3 students per class for maximum hands-on attention and personalized guidance.
- Live Online Cohorts: Small groups with a maximum of 6 students with real-time screen sharing and collaborative debugging.
- Zero Homework Anxiety: All required building, testing, and concept mastery happen in class.
Why Learn Train a Smart AI Model?
Artificial Intelligence and Machine Learning are transforming every industry on the planet. Yet young people are frequently treated only as consumers of algorithmic feeds and generative AI tools, without understanding how these systems actually function under the hood.
Train a Smart AI Model demystifies machine learning through hands-on empirical experimentation. Using Google Teachable Machine and custom web runners, kids aged 8 to 13 train computer vision and sound recognition neural models directly using their laptops’ webcams and microphones. They discover the difference between traditional rule-based coding and data-driven machine learning, stress-test models against background noise and lighting bias, and author an official AI Model Card. They emerge as informed, critical, and ethical AI creators.
Skills Your Child Will Practise
- Rules vs Data Paradigms: Understanding how machine learning learns patterns from examples rather than static if-statements.
- Webcam Computer Vision: Collecting balanced image datasets, training multi-class vision models, and testing live confidence meters.
- Audio Spectrograms & Ambient Calibration: Visualizing sound frequencies and calibrating background noise for robust triggers.
- Bias & Edge Case Auditing: Diagnosing background bias, false positives, lighting distortions, and spurious correlations.
- Responsible AI & Model Cards: Authoring documentation on training data provenance, operational limits, and fairness.
The Course Journey
The course has six live one-hour meetings. Chapters are learning units: a chapter may span meetings, with time for practice, questions and saving work. Optional extensions depend on readiness and available time.
| Module | Focus & Learning Objectives | Tangible Evidence of Learning |
|---|---|---|
| Chapter 01 | Teaching the computer to see Understand rules vs data; collect webcam samples; train a visual classifier and save a .tm file. | Demonstrate a live 3-class visual model and reopen a saved .tm project. |
| Chapter 02 | The sneaky background Investigate background bias and spurious correlation; stress-test with angles, distances, and lighting; retrain for robustness. | Document an edge-case failure and demonstrate the retrained model succeeding. |
| Chapter 03 | Teaching the computer to listen Explore audio spectrograms; calibrate ambient background noise; train a 2-sound/speech recognition model. | Demonstrate live audio triggers that ignore room noise. |
| Chapter 04 | The voice challenge Test pitch, speed, and confuser words; identify false positives/negatives; add a negative class and tune thresholds. | Demonstrate command recognition across vocal variations while rejecting rhyming words. |
| Chapter 05 | The interactive AI controller Export a trained model; connect it to the interactive web runner; tune confidence thresholds and debounce. | Operate an interactive web preview driven by webcam gestures or voice commands. |
| Chapter 06 | The AI audit and showcase Conduct a peer audit; evaluate real-world algorithmic bias; complete an AI Model Card; present your project. | Present your running AI model with its completed Model Card and explain one limitation. |
Course Details at a Glance
Target Age & Readiness
Recommended Age: 8+
Prerequisites: Can use a mouse or trackpad and type short labels. No coding experience needed.
Class Format & Schedule
Duration: 6 weekly sessions × 60 minutes
Tuition: See the current total price for your chosen class on the registration page before booking.
Equipment & Tools
Equipment: A laptop with webcam, microphone, and a modern browser (Chrome, Edge, or Safari).
Work Preservation: All projects are saved locally on the student's computer.
Takeaway Creation
Project Outcome: A custom-trained AI vision and sound controller with a tested model card.
Ready to showcase to family, teachers, and friends.
Frequently Asked Questions for Parents
Got questions about enrolling, class structure, or student readiness? Here are the most common questions Ottawa parents ask us:
Where are in-person classes held?
In Stittsville, Ottawa, convenient for families in Stittsville, Kanata, Richmond, Carp, and West Ottawa.
Is there an online class option?
Yes, live online sessions allow kids to train webcam and microphone models from home with real-time mentor guidance.
What age group is this designed for?
Ages 8 to 13. No math prerequisites or coding background needed; students use interactive visual machine learning tools.
What tools and equipment are used?
A laptop with webcam, microphone, and a modern browser (Chrome, Edge, or Safari). We use Google Teachable Machine and custom web runners. No accounts or personal data storage required.
What core concepts will children understand?
Rules vs data-driven programming, webcam sample collection, background bias and spurious correlation, audio spectrograms, ambient noise calibration, confidence thresholds, and ethical AI auditing.
What is the final project?
A custom-trained multimodal AI controller operating a live web game, accompanied by a documented AI Model Card detailing capabilities, blind spots, and fairness testing.
Ready for Your Child to Become a Creator?
Choose from currently available dates and teaching modes. Groups have a maximum of 3 learners in person or 6 online.