Open learning resources

From data to AI to its responsible deployment.

A guide to practical data science, machine learning foundations, and responsible AI, organized by learner goals.

Find a starting point Browse topics Try the Readiness Arcade

Start here

What do you want to do next?

Before ML

Warm up with the Readiness Arcade

Use a self-diagnostic and five short practice arcades to refresh the math, Python, and scientific habits that make machine learning easier to enter.

Next: Open the diagnostic and choose only the arcades you need.

Open the arcade ↓

Practical Data Science

I’m new to data science

Begin with Practical Data Science: Python, notebooks, tabular data, cleaning, visualization, statistics, uncertainty, and first ML models.

Next: Open CSCI 1109 and start with Orientation & Python Foundations.

Start here ↓

Machine Learning Foundations

I want to understand machine-learning models

Use CSCI 3151 to move from learning paradigms and evaluation to kernels, neural networks, representation learning, and transformers.

Next: Start with ML Paradigms, then use the topic index when you need a specific concept.

Start here ↓

Responsible AI

I make decisions about AI systems

Start with Responsible AI by Design: problem framing, ethical principles, fairness, explainability, governance, privacy, law, procurement, and culture.

Next: Use each module as a checklist for decisions, vendors, policies, or leadership conversations.

Start here ↓

Practical Data Science + Machine Learning Foundations + Responsible AI

I work in health, public service, or civic technology

Combine data literacy from CSCI 1109, model literacy from CSCI 3151, and deployment/governance literacy from Responsible AI.

Next: Follow the full arc: evidence → model behaviour → governance and public accountability.

Start here ↓

Practical Data Science + Machine Learning Foundations

I need a bridge into the math

Use the math, descriptive statistics, probability, and linear-algebra portions of CSCI 1109 now.

Next: Look for vectors, matrices, descriptive statistics, distributions, sampling, and bootstrapping.

Start here ↓

Courses, catalogs, and warmups

Four ways into data, ML, and responsible AI

These are each complete, individual courses, made free and open.

Static arcade

Warmup

Readiness Arcade

Pre-course

A compact refresher for the math, Python, and evidence habits that students need before learning machine learning.

Best for: incoming ML students, self-study learners, and anyone who wants to self-assess before learning more deeply.

Start-screen self-assessment for choosing a route.
Five arcades: linear algebra, calculus, probability, NumPy, and evidence.
Notebook practice ending with a brief gateway.

5 arcades

1 notebook gym

1 final boss

Open arcade View map

Course

CSCI 1109

Practical Data Science

Current public course · Introductory

Start here to learn how real-world questions become datasets, analyses, visualizations, models, and decisions.

Best for: Learners starting with data science, Python, tabular data, and evidence-based decisions.

Work with Python notebooks, pandas, NumPy, Matplotlib, scikit-learn, and NetworkX.
Move from messy data to visualization, inference, causal thinking, and first machine-learning models.
Build habits around cleaning, testing, evaluation, fairness, privacy, and communication.

60+ modules

11 clusters

Intro → core level

Open course View map

Course

CSCI 3151

Foundations of Machine Learning

Current public course · Core undergraduate

Use this path to understand how machine-learning models are trained, evaluated, regularized, generalized, and connected to modern deep-learning architectures.

Best for: Students and self-directed learners who know some programming and want the conceptual machinery behind modern ML.

Develop a vocabulary for supervised, unsupervised, semi-supervised, and self-supervised learning.
Study likelihood, EM, model evaluation, bias–variance, regularization, optimization, kernels, and SVMs.
Connect foundational ideas to neural networks, CNNs, RNNs, transformers, embeddings, and representation learning.

60+ modules

4 assignments

Core ML level

Open course View map

Professional course

Microcredential

Responsible AI by Design

Public course materials · Professional / leadership

Use this path to ask better questions before AI systems are procured, adopted, governed, or communicated to affected people.

Best for: Business leaders, managers, product owners, procurement leads, compliance professionals, and public-sector decision makers.

Translate responsible AI principles into practical questions for leadership, procurement, governance, and deployment.
Evaluate bias, fairness, explainability, accountability, privacy, legal readiness, vendor risk, and organizational culture.
Build toward an AI Leadership Action Plan with owners, evidence artifacts, monitoring, escalation, and communication steps.

10 modules

~12 hours

Leaders audience

Open materials View map

Search topics

Search by keyword

Practical tools

Notebooks, Python, and data tables

For learners who need the everyday mechanics of Jupyter, Python, tabular data, reproducible cells, and good data habits.

Python Jupyter notebooks conda pandas CSV JSON dataframes tidy data data cleaning

module CSCI 1109 M02 — Python setup: notebooks, VS Code, conda Use this for the basic computing environment and where notebook code is running. module CSCI 1109 M07 — Series vs DataFrame; tidy data Use this for pandas tables, rows, columns, indices, and tidy-data habits. static page Readiness Arcade setup Use this short setup path before opening the notebook gym.

Math readiness

Vectors, matrices, shape, and dot products

To refresh linear algebra: rows, columns, dot products, matrix-vector multiplication, projections, and shape debugging.

vectors matrices linear algebra dot products shape projection matrix-vector multiplication bias column standardization

notebook Readiness Arcade — Arcade 1 notebook gym Practice dot products, norms, projections, matrix-vector shapes, bias columns, and standardization. module CSCI 1109 M19 — Vectors, matrices, broadcasting; linear ops Open the practical data-science module where matrix and vector operations first appear. module CSCI 3151 M26 — Linear dimensionality reduction & PCA Use this when linear algebra starts driving representation and compression.

Math readiness

Calculus, local change, and gradients

For learners who remember derivatives vaguely but want to apply it to machine learning.

calculus derivatives gradients partial derivatives directional derivative finite difference sigmoid optimization gradient descent

notebook Readiness Arcade — Arcade 2 notebook gym Refresh derivatives, gradients, tangent lines, finite differences, and local-change explanations. module CSCI 3151 M17 — Optimization basics: gradient descent & variants Connect derivatives to learning updates, loss surfaces, and optimization stability. module CSCI 3151 M31 — Backpropagation & automatic differentiation Use this when gradients move through computational graphs and neural networks.

Uncertainty

Probability, base rates, and distributions

For learners who need to reason clearly about events, conditional probability, Bayes’ rule, sample variation, and what rates actually mean.

probability base rates Bayes Bayes' rule conditional probability distributions variance uncertainty sampling bootstrap

notebook Readiness Arcade — Arcade 3 notebook gym Work through base-rate intuition, conditional tables, Bayes’ rule, expected value, variance, and split base rates. module CSCI 1109 M22 — Descriptive statistics & distributions Use this for distributions, descriptive summaries, and introductory probability. module CSCI 3151 M08 — Maximum likelihood estimation Use this when probability becomes an objective for learning model parameters.

Computing

NumPy, vectorization, and small algorithms

For learners who can read Python: arrays, masks, broadcasting, distances, and simple KNN from scratch.

NumPy vectorization broadcasting boolean masks one-hot Euclidean distance KNN nearest neighbours top-k

notebook Readiness Arcade — Arcade 4 notebook gym Practice vectorized operations, masking, one-hot encoding, distances, KNN votes, and missing-value handling. notebook CSCI 1109 M20/A3 — NumPy and vector/matrix practice notebook Open the assignment notebook used for vectors, bootstrapping, and visualization practice. module CSCI 1109 M41 — k-NN intuition; distance & scaling Connect vectorized distances to a concrete classifier.

Evaluation

Validation, leakage, metrics, and evidence

For learners who need to distinguish training, validation, and testing, avoid leakage, and explain model claims responsibly.

validation test set train set leakage metrics confusion matrix accuracy precision recall thresholds

module CSCI 3151 M12 — Evaluation metrics for regression & classification Use this for accuracy, precision, recall, F1, ROC–AUC, and metric trade-offs. module CSCI 3151 M13 — Train/validation/test splits & cross-validation Use this for honest evaluation design, split strategy, cross-validation, and leakage. module CSCI 1109 M40 — Baselines & leakage traps Use this for practical baseline-first thinking and leakage red flags.

Communication

Plots, evidence, and communication

For learners who need to communicate results clearly: plots, labelled comparisons, summary traps, and concise evidence statements.

plots visualization scatter plot evidence statement communication Anscombe summaries figures storytelling

module CSCI 1109 M25 — Visualization & Storytelling Start here for chart grammar, visual encodings, and choosing plots for decisions. module CSCI 1109 M29 — From EDA to narrative Use this for turning exploratory analysis into a clear, annotated result. notebook Readiness Arcade — Arcade 5 notebook gym Practice scatter plots, confusion matrices, validation curves, and short evidence statements.

Machine learning foundations

Supervised, unsupervised, and learning paradigms

For learners of basic machine learning paradigms.

supervised unsupervised self-supervised semi-supervised classification regression clustering paradigms

module CSCI 3151 M02 — Learning settings Use this for supervised, unsupervised, self-supervised, and semi-supervised distinctions. module CSCI 3151 M04 — Supervised learning pipeline & loss functions Use this when labels, losses, models, and metrics need to fit together. module CSCI 1109 M38 — What is ML? tasks; train/val/test splits Use this for a gentler practical-data-science entrance into machine learning.

Model fitting

Loss, optimization, regularization, and capacity

For learners moving from basic predictions to slightly more 'under the hood' detail.'

loss objective regularization L1 L2 ridge lasso capacity overfitting underfitting

module CSCI 3151 M14 — Bias–variance tradeoff & learning curves Use this for diagnosing underfitting, overfitting, and generalization gaps. module CSCI 3151 M16 — Regularization: L2, L1, and beyond Use this for shrinkage, sparsity, weight decay, and tuning regularization strength. module CSCI 3151 M17 — Optimization basics Use this for gradient descent and the mechanics of training.

Prediction

Classification, KNN, logistic regression, and thresholds

For learners comparing local-neighbour prediction, probability-based classification, and threshold decisions under different error costs.

KNN nearest neighbours logistic regression classification sigmoid decision threshold precision recall false positives false negatives

module CSCI 1109 M41 — k-NN intuition; distance & scaling Use this for a concrete distance-based classifier and why scaling matters. module CSCI 3151 M05 — From regression to logistic regression Use this for probabilistic binary classification and why logistic loss matters. module CSCI 1109 M47 — Logistic regression & error metrics Use this for confusion matrices, thresholds, precision, recall, and F1.

Representation

Feature spaces, dimensionality reduction, and representations

For learners of features and representations, how high-dimensional data can be visualized, and what learned embeddings are doing.

feature spaces feature engineering PCA t-SNE UMAP dimensionality reduction embeddings representations feature maps

module CSCI 3151 M23 — Feature engineering principles Use this for feature design, preprocessing pipelines, leakage, and proxy risks. module CSCI 3151 M26 — Linear dimensionality reduction & PCA Use this for PCA, explained variance, and compression. module CSCI 3151 M40 — Principles of representation learning Use this for learned feature spaces, invariances, and objectives shaping representations.

Deep learning

Neural networks, backpropagation, and deep training stability

For learners moving from small models to multilayer networks, gradient flow, and the practical issues one encounters in ML.

neural networks MLP feedforward backpropagation autodiff gradient flow vanishing gradients exploding gradients dropout batchnorm

module CSCI 3151 M30 — Feedforward neural networks Use this for the structure of MLPs, nonlinear activations, capacity, and basic tuning. module CSCI 3151 M31 — Backpropagation & automatic differentiation Use this for the chain rule, computation graphs, and PyTorch autograd. module CSCI 3151 M38 — Dropout, batch normalization, and training tricks Use this for regularization and training recipes in deeper networks.

Modern architectures

CNNs, RNNs, attention, and transformers

For learners who want to navigate the architectures behind image, sequence, and language models.

CNN convolutions ResNet RNN LSTM GRU attention self-attention transformers sequence modelling

module CSCI 3151 M44 — Convolutions, pooling, and CNN architectures Use this for image models, kernels, pooling, channels, and receptive fields. module CSCI 3151 M48 — Sequence modelling & vanilla RNNs Use this for ordered data, hidden states, and recurrent baselines. module CSCI 3151 M52 — Attention mechanisms & self-attention Use this for queries, keys, values, weighted retrieval, and the transformer setup.

Responsibility

Fairness, accountability, governance, and responsible deployment

For learners and leaders who want to deploy AI and ML responsibly in the real world.

fairness bias accountability governance Responsible AI deployment stakeholders public interest risk model cards

module Responsible AI Module 3 — Bias and Fairness Use this for bias sources, fairness metrics, mitigation, and monitoring. module Responsible AI Module 5 — Accountability and Governance Use this for RACI, assurance evidence, governance functions, and lifecycle accountability. module CSCI 1109 M57 — Algorithmic bias & fairness frames Use this for a practical-data-science version of subgroup error and fairness frames.

AI governance

Privacy, law, procurement, and organizational controls

For learners evaluating or buying AI systems: what evidence to ask for, what legal context matters, and how to turn responsible AI into operational controls.

privacy data ethics legal regulation EU AI Act Canada procurement vendor risk contracts security

module Responsible AI Module 6 — Privacy and Data Ethics Use this for re-identification, inference risk, differential privacy, and data ethics. module Responsible AI Module 7 — Legal and Regulatory Landscape Use this for Canada, the EU AI Act, sector rules, and compliance-readiness questions. module Responsible AI Module 8 — Responsible AI Procurement Use this for evidence packs, contracts, vendor controls, and lifecycle procurement.

Trust and audit

Explainability, transparency, and model evidence

For learners who need to communicate what a model is doing internally.

explainability interpretability transparency SHAP LIME counterfactuals black box audit recourse

module Responsible AI Module 4 — Transparency and Explainability Use this for interpretable models, XAI tools, counterfactuals, and explanation limits. module CSCI 3151 M23 — Feature engineering principles Use this for the technical side of feature choices, pipelines, and leakage. module Responsible AI Module 5 — Assurance evidence Use this for model cards, system cards, logs, monitoring, and audit evidence.

Use or reuse

For instructors and teams

Assign, adapt, or reuse these materials.

Instructors

Assign a course, cluster, or module

Use the course cards and maps to choose a full sequence or a smaller unit.

Choose a resource ↓

Teams

Use this for AI literacy and governance

Use Responsible AI by Design for fairness, privacy, procurement, monitoring, and accountability conversations.

Open Responsible AI ↓

Reuse terms

Free to reuse with attribution

Site code is GPLv3. Course materials are CC BY-NC-SA 4.0 unless otherwise noted. Third-party materials keep their own licenses.

See reuse notes ↓

Canada AI literacy alignment Optional crosswalk

These resources support data fluency, model evidence, critical verification, governance, privacy, fairness, and accountable deployment.

Data literacy and privacy

CSCI 1109 covers data cleaning, visualization, statistics, and uncertainty. Responsible AI adds privacy, data ethics, and governance controls.

Critical thinking and verification

The Readiness Arcade, CSCI 1109, and CSCI 3151 emphasize baselines, leakage, validation, metrics, and evidence statements.

Foundations of AI knowledge

CSCI 3151 builds from learning paradigms and evaluation to optimization, neural networks, attention, transformers, and representations.

Ethical awareness and responsible use

Responsible AI by Design covers fairness, explainability, accountability, privacy, legal readiness, procurement, and vendor risk.

Open Education repositories

This hub has been submitted for indexing in several Open Education repositories. Links will appear here when available.

From data to AI to its responsible deployment.

What do you want to do next?

Warm up with the Readiness Arcade

I’m new to data science

I want to understand machine-learning models

I make decisions about AI systems

I work in health, public service, or civic technology

I need a bridge into the math

Four ways into data, ML, and responsible AI

Readiness Arcade

Practical Data Science

Foundations of Machine Learning

Responsible AI by Design

Search by keyword

Notebooks, Python, and data tables

Vectors, matrices, shape, and dot products

Calculus, local change, and gradients

Probability, base rates, and distributions

NumPy, vectorization, and small algorithms

Validation, leakage, metrics, and evidence

Plots, evidence, and communication

Supervised, unsupervised, and learning paradigms

Loss, optimization, regularization, and capacity

Classification, KNN, logistic regression, and thresholds

Feature spaces, dimensionality reduction, and representations

Neural networks, backpropagation, and deep training stability

CNNs, RNNs, attention, and transformers

Fairness, accountability, governance, and responsible deployment

Privacy, law, procurement, and organizational controls

Explainability, transparency, and model evidence

For instructors and teams

Assign a course, cluster, or module

Use this for AI literacy and governance

Free to reuse with attribution

Data literacy and privacy

Critical thinking and verification

Foundations of AI knowledge

Ethical awareness and responsible use

Open Education repositories

Full coverage map

Pre-course gym

Start-screen diagnostic

Vectors, matrices, dot products, and shape

Calculus and local change

Probability, base rates, and distributions

Python, NumPy, vectorization, and small algorithms

Plots, evidence, and communication

Final KNN evidence memo

Practical Data Science

Orientation & Python Foundations

Tabular Data with pandas

Cleaning & Preparation (Pipelines)

Math & Descriptive Stats

Visualization & Storytelling

Probability, Testing & Causal Thinking

Intro to ML: Supervised Learning

Regression & Classification

Unsupervised & Structure in Data

Ethics, Privacy & Responsible DS

Communication & Synthesis

Machine Learning Foundations

Intro to ML Paradigms

Supervised & Unsupervised Learning

Maximum Likelihood Estimation & EM

Model Evaluation & Bias–Variance

Regularization & Optimization Basics

Kernel Methods

Feature Engineering & Data Augmentation

Dimensionality Reduction

Neural Networks & Backpropagation

Training Stability

Generalization Techniques

Representation Learning & Embeddings

Convolutional Neural Networks

Recurrent Neural Networks

Transformers

Self-Supervised & Semi-Supervised Learning

Course Review & Synthesis

Responsible AI

AI Foundations and Ethical Principles