School of Data Science
University of Virginia
Course Overview
Course Calendar
Course Policies
View the Project on GitHub thomasgstewart/machine-learning-1-fall-2023
The final project presentations will be Saturday, December 9, 2023 from 9:00AM-12:00PM.
| Title | First Submission Due Date | Resubmission Due Date |
|---|---|---|
| 1. Student Profile (not graded) | 2023-08-28 | Not available |
| 2. Linear independence | 2023-09-04 | |
| 3. Appendix A Exercises | 2023-09-08 | |
| 4. Nonlinear | 2023-10-02 | |
| 5. PCA | 2023-10-18 | |
| 6. Constrained kernel | 2023-10-18 | |
| 7. Optimism | 2023-12-05 | |
| 8. Ordinal regression model | 2023-12-05 | |
| 9. Logistic regression model | 2023-12-05 |
Material in monospaced font are in the course Teams site. The letter H denotes sections in Regression Modeling Strategies; the letter C denotes sections in Plane Answers.
| Class date | Topic | Material |
|---|---|---|
| Data types | 01-data-types |
|
| Course overview | ||
| 2023-08-28 | Linear Algebra Review | C Appendix A, C Appendix B, C 1 |
| → Vector space | ||
| → Vector subspace | ||
| → Matrix as a function | ||
| → Span | ||
| → Column space | ||
| → Linear dependence | ||
| → Basis | ||
| → Rank | ||
| →→ Uniqueness | ||
| → Sum of subspaces | ||
| →→ Uniqueness | ||
| → Orthogonal vectors | ||
| → Orthogonal basis | ||
| → Orthonormal basis | ||
| → Gram-Schmidt | ||
| → Orthogonal complement | ||
| →→ Decomposition of vector space into subspace and orthogonal complement | ||
| Linear Regression | ||
| Model formulation | Slides in Teams | |
| Bayesian vs MLE | Slides in Teams | |
| Interactions | Slides in Teams | |
| Nonlinearity | Slides in Teams | |
| Measures of model performance | ||
| → Discrimination | ||
| → Calibration | ||
| → Optimism | ||
| Measures of model performance | ||
| Carrying capacity of data | ||
| Model complexity | ||
| Strategies for right-sizing the model complexity | ||
| → regularization (LASSO, ridge, Bayesian) | ||
| → constraints (principle components, monotonicity) | ||
| Logistic Regression | ||
| Model formulation | ||
| Bayesian vs MLE | ||
| Interactions | ||
| Nonlinearity | ||
| Measures of model performance | ||
| → Discrimination | ||
| → Calibration | ||
| → Optimism | ||
| Measures of model performance | ||
| Carrying capacity of data | ||
| Model complexity | ||
| Strategies for right-sizing the model complexity | ||
| → regularization (LASSO, ridge, Bayesian) | ||
| → constraints (principle components, monotonicity) | ||
| Ordinal Regression | ||
| Model formulation | ||
| Bayesian vs MLE | ||
| Interactions | ||
| Nonlinearity | ||
| Measures of model performance | ||
| → Discrimination | ||
| → Calibration | ||
| → Optimism | ||
| Measures of model performance | ||
| Carrying capacity of data | ||
| Model complexity | ||
| Strategies for right-sizing the model complexity | ||
| → regularization (LASSO, ridge, Bayesian) | ||
| → constraints (principle components, monotonicity) | ||
| Hazard Regression | ||
| Model formulation | ||
| Bayesian vs MLE | ||
| Interactions | ||
| Nonlinearity | ||
| Measures of model performance | ||
| → Discrimination | ||
| → Calibration | ||
| → Optimism | ||
| Measures of model performance | ||
| Carrying capacity of data | ||
| Model complexity | ||
| Strategies for right-sizing the model complexity | ||
| → regularization (LASSO, ridge, Bayesian) | ||
| → constraints (principle components, monotonicity) | ||
| Random Forest | ||
| Model formulation | ||
| Bayesian vs MLE | ||
| Interactions | ||
| Nonlinearity | ||
| Measures of model performance | ||
| → Discrimination | ||
| → Calibration | ||
| → Optimism | ||
| Measures of model performance | ||
| Carrying capacity of data | ||
| Model complexity | ||
| Strategies for right-sizing the model complexity | ||
| → regularization (LASSO, ridge, Bayesian) | ||
| → constraints (principle components, monotonicity) |