Google Teachable Machine
Topics: Supervised classification (image, sound, pose); collecting/labeling training data; training vs. testing a model; model confidence.
Prerequisites: None; basic webcam/browser use.
Code.org – AI & Machine Learning Modules
Topics: AI vs. traditional programming; how AI is trained on data; bias in training data and AI fairness; intro to chatbots/LLMs.
Prerequisites: None; grade-appropriate reading level.
MIT Day of AI
Topics: AI ethics and societal impact; pattern recognition and data-driven decisions; algorithmic bias and data quality.
Prerequisites: None; designed for classroom/family use with adult facilitation.
Machine Learning for Kids
Topics: Full ML loop (collect data, train, test, evaluate, iterate) for text/image/number/sound classifiers; applying models inside Scratch 3; optional Python track.
Prerequisites: Basic Scratch familiarity helpful; free IBM Cloud account (adult sign-up may be needed).
Scratch (with AI/ML Extensions)
Topics: Event-driven programming logic (loops, conditionals, variables); integrating trained ML models into interactive projects; sensor input → model → output flow.
Prerequisites: None to start; prior Code.org exposure helpful.
Khan Academy – AI for Education
Topics: How AI/LLMs work (tokens, prediction); what machine learning is; AI bias, ethics, and responsible use.
Prerequisites: Basic algebra recommended; no coding required.
Google Machine Learning Crash Course
Topics: Linear/logistic regression, loss, gradient descent; classification metrics (precision, recall, AUC); neural networks, embeddings, LLMs, AutoML; Responsible AI/fairness.
Prerequisites: Algebra-level math; Python, NumPy, and pandas familiarity recommended.
Kaggle Learn
Topics: Python and Pandas; decision trees/random forests; pipelines and cross-validation; intro deep learning, CNNs, NLP, time series; AI ethics.
Prerequisites: Basic Python helpful (taught from scratch if needed).
Coursera – Machine Learning Specialization
Topics: Supervised learning (linear/logistic regression, neural networks); decision trees/ensembles; unsupervised learning (k-means, anomaly detection); recommender systems; reinforcement learning basics.
Prerequisites: Basic Python; high-school-level algebra.
edX (MIT 6.86x / Harvard CS50 AI)
Topics: Linear classifiers, SVMs, kernels, neural networks; search algorithms, knowledge representation, Bayesian networks; NLP.
Prerequisites: Varies by course; intro tracks need basic programming, advanced tracks need calculus/linear algebra/probability.
DeepLearning.AI
Topics: Neural network foundations, hyperparameter tuning, regularization; CNNs for computer vision; RNNs/LSTMs, attention, Transformers; generative AI (RAG, fine-tuning, diffusion).
Prerequisites: Intermediate Python; linear algebra, calculus, and probability; Coursera ML Specialization recommended first.
fast.ai – Practical Deep Learning for Coders
Topics: Applied deep learning (vision, NLP, tabular, collaborative filtering) with PyTorch/fastai; transfer learning; SGD, backprop, optimizers built from scratch; model deployment; Part 2: diffusion models, research papers.
Prerequisites: 1+ year coding experience (Python preferred); high-school math (rest taught in-course).
MIT OpenCourseWare – AI & Machine Learning
Topics: 6.036/6.390 Intro to Machine Learning; 6.S191/6.7960 Deep Learning; supervised/unsupervised learning, CNNs, RNNs, generative models, reinforcement learning.
Prerequisites: Multivariable calculus, linear algebra, probability/statistics; solid Python programming.
Stanford Online – CS229 / CS231n
Topics: CS229: statistical pattern recognition, GLMs, SVMs, kernel methods, learning theory, reinforcement learning/MDPs. CS231n: CNN architectures, training/fine-tuning for computer vision.
Prerequisites: Python/NumPy proficiency; probability theory; multivariable calculus and linear algebra (Stanford CS106/109/MATH51 equivalents).