STEM / applied
Machine learning for statistics · Graduate · Math
Learning objectives from the Machine learning for statistics syllabus, grouped by unit. Click an objective for study materials.
Topics typically covered
Click a topic for the full text and related unit practice.
Graduate Machine learning for statistics — outline derived from course README sections and typical US statistics syllabi (OpenIntro / standard OER where applicable).
Statistical learning foundations
- Loss functions — Loss functions
- Empirical risk minimization — Empirical risk minimization
- Bias-variance — Bias-variance
- Generalization error — Generalization error
- VC dimension — VC dimension
- Complexity control (intro) — Complexity control (intro)
- Regularization paths — Regularization paths
- Model selection — Model selection
- Cross-validation theory — Cross-validation theory
- Practice — Practice
Modern methods
- Ensemble learning: boosting — Ensemble learning: boosting
- Bagging — Bagging
- Kernel methods and SVMs — Kernel methods and SVMs
- Neural networks from a statistical view — Neural networks from a statistical view
- Unsupervised learning: clustering — Unsupervised learning: clustering
- Embeddings — Embeddings
- Feature selection and sparsity — Feature selection and sparsity
Evaluation and deployment
- Proper scoring rules — Proper scoring rules
- Calibration and fairness metrics — Calibration and fairness metrics
- Interpretability: SHAP — Interpretability: SHAP
- LIME (overview) — LIME (overview)
- Statistical inference after model selection (intro) — Statistical inference after model selection (intro)
Learning objectives
Click an objective for study materials.
Statistical learning foundations
- Loss functions — Loss functions
- Empirical risk minimization — Empirical risk minimization
- Bias-variance — Bias-variance
- Generalization error — Generalization error
- VC dimension — VC dimension
- Complexity control (intro) — Complexity control (intro)
- Regularization paths — Regularization paths
- Model selection — Model selection
Modern methods
- Ensemble learning: boosting — Ensemble learning: boosting
- Bagging — Bagging
- Kernel methods and SVMs — Kernel methods and SVMs
- Neural networks from a statistical view — Neural networks from a statistical view
- Unsupervised learning: clustering — Unsupervised learning: clustering
- Embeddings — Embeddings
- Feature selection and sparsity — Feature selection and sparsity
Evaluation and deployment
- Proper scoring rules — Proper scoring rules
- Calibration and fairness metrics — Calibration and fairness metrics
- Interpretability: SHAP — Interpretability: SHAP
- LIME (overview) — LIME (overview)
- Statistical inference after model selection (intro) — Statistical inference after model selection (intro)
Definitions and structure
- Loss functions — Loss functions
- Empirical risk minimization — Empirical risk minimization
- Bias-variance — Bias-variance
- Generalization error — Generalization error
- VC dimension — VC dimension
Proofs and reasoning
- Complexity control (intro) — Complexity control (intro)
- Regularization paths — Regularization paths
- Model selection — Model selection
- Cross-validation theory — Cross-validation theory
- Practice — Practice
Abstraction and generalization
- Ensemble learning: boosting — Ensemble learning: boosting
- Bagging — Bagging
- Kernel methods and SVMs — Kernel methods and SVMs
- Neural networks from a statistical view — Neural networks from a statistical view
- Unsupervised learning: clustering — Unsupervised learning: clustering
Modeling and computation
- Apply loss functions in engineering contexts — Apply loss functions in engineering contexts
- Apply empirical risk minimization in engineering contexts — Apply empirical risk minimization in engineering contexts
- Apply bias-variance in engineering contexts — Apply bias-variance in engineering contexts
- Apply generalization error in engineering contexts — Apply generalization error in engineering contexts
- Apply vc dimension in engineering contexts — Apply vc dimension in engineering contexts
Data and technology
- Use software to explore machine learning for statistics problems nume... — Use software to explore machine learning for statistics problems numerically
- Interpret computational results against analytic predictions — Interpret computational results against analytic predictions
- Build spreadsheets or scripts for routine calculations — Build spreadsheets or scripts for routine calculations
- Visualize functions, fields, or datasets tied to course topics — Visualize functions, fields, or datasets tied to course topics
- Connect course methods to lab, industry, or research workflows — Connect course methods to lab, industry, or research workflows
Problem-solving practice
- Complexity control (intro) — Complexity control (intro)
- Regularization paths — Regularization paths
- Model selection — Model selection
- Cross-validation theory — Cross-validation theory
- Practice — Practice
Multi-Unit Problems
Course-level sets that combine skills across study units (coming soon).
Browse Multi-Unit ProblemsWhat each unit includes
Open a unit below for full materials. Typical resources:
- Study guide
- Exam Strategy
- Common Mistakes
- Worksheets
- Word problems
- Mixed Practice
- Multi-Unit Problems
- Review
- Practice test
- Answer key
Study units
Each unit includes a study guide, worksheets, review, practice test, and answer key. One unit is free; subscribe for the full class.
- Statistical learning foundations
Loss functions
Coming soon - Modern methods
Ensemble learning: boosting
Coming soon - Evaluation and deployment
Proper scoring rules
Coming soon
Pricing calculator
Choose materials, tutoring, or both — or book a single session as needed. Customize your plan on the subscribe page.
$1,162 · Machine learning for statistics · 18 tutoring hrs
Study guides, worksheets, reviews, practice tests, and answer keys for 1 class. 18 tutoring hours (1 hr / week · semester). Bundle discount applied vs buying separately. Pay in full via Zelle.