ECON526: Graduate Quantitative Economics and Datascience
Getting Started
- Schedule: lecture dates, problem set deadlines, and the midterm
- Exams and Scheduling: self-scheduled assessments and ORCA reservations
- Cold calling and participation: classroom questions, random selection, and participation grading
- Python setup: install Python, Jupyter, and VS Code
- Canvas submissions: problem sets are submitted as
ipynband exportedhtmlfiles - Notebooks repository: every deck and problem set as a runnable notebook
Slides
| Lecture | Formats |
|---|---|
|
Foundations of Numerical Linear Algebra Vectors, matrices, and linear systems in NumPy and SciPy, LU and triangular solves, orthogonal matrices, eigendecompositions, spectral radius, and conditioning. |
HTML · PDF · Notebook |
|
Linear Algebra and Dynamics Matrix powers and eigendecompositions for linear difference equations, steady states as fixed points, the Neumann series, unemployment dynamics, PageRank, present discounted values, and an optional section on linearizing nonlinear dynamics. |
HTML · PDF · Notebook |
|
Optimization, Positive Definiteness, and Least Squares Curvature and definiteness in optimization, positive definite matrices and the Cholesky, simulating correlated Gaussian shocks from a covariance matrix, then least squares through the normal equations, collinearity and identification, and regularization. |
HTML · PDF · Notebook |
|
Probability and Uncertainty An optional review of probability spaces, then discrete distributions and expectations, the law of large numbers and central limit theorem, joint, marginal, and conditional distributions, Simpson’s paradox and Bayes’ law, and conditional and iterated expectations. |
HTML · PDF · Notebook |
|
Classification Logistic and multinomial regression as classifiers, decision rules and their evaluation, coding occupations from job titles, and the measurement error a classifier introduces into a regression. |
HTML · PDF · Notebook |
|
Clustering and Discrete Latent Variables Latent variables and unsupervised learning, with k-means clustering to uncover discrete latent groups and tools for choosing the number of clusters. |
HTML · PDF · Notebook |
|
Dimension Reduction and Continuous Latent Variables Continuous latent variables with the singular value decomposition, principal components, and auto-encoders, with an optional factor model of a portfolio. |
HTML · PDF · Notebook |
|
Text Classification Word counts, TF-IDF, supervised sentiment and topic classifiers, and coverage regressions without LLMs or pretrained embeddings. |
HTML · PDF · Notebook |
|
Natural Language Processing and Large Language Models Classify news with embeddings and LLM prompts, then explain tokens, latent representations, and sequential token prediction. |
HTML · PDF · Notebook |
|
Open-Weight LLMs on Your Machine Install Ollama, run open models for chat and embeddings locally, and call them from Python through OpenAI-compatible APIs. |
HTML · PDF · Notebook |
|
Handwriting and Digitization Download and call handwriting models, distinguish training from evaluation, and compare local OCR, general vision-language models, and corpus-specific fine-tunes. |
HTML · PDF · Notebook |
|
Image Classification Turn satellite images into land-use variables with a local vision-language model and a downloaded classifier, and compare them on the same chips. |
HTML · PDF · Notebook |
|
Spatial Data for Economics Geographic data, spatial joins, commuting shares, and a commuting gravity regression. |
HTML · PDF · Notebook |
|
Stochastic Processes, Markov Chains, and Expectations Stochastic processes, Markov chains, stationary distributions, and conditional expectations. |
HTML · PDF · Notebook |
No matching items
Problem Sets
| Problem Set | Formats |
|---|---|
|
ECON526: Problem Set 0 Minimize the function \(x^2 + y^2 +z^2\) subject to |
HTML · PDF · Notebook |
|
ECON526: Problem Set 1 Install Python (with Jupyter and VS Code) |
HTML · PDF · Notebook |
|
ECON526: Problem Set 2 Use the following packages and imports |
HTML · PDF · Notebook |
|
ECON526: Problem Set 3 Feel free to use the following packages (and we have added a few convenience imports) |
HTML · PDF · Notebook |
|
ECON526: Midterm 1 Practice Ungraded practice problems for Midterm 1. |
HTML · PDF · Notebook |
|
ECON526: Problem Set 4 Feel free to use the following packages (and we have added a few convenience imports) |
HTML · PDF · Notebook |
|
ECON526: Problem Set 5 Feel free to use the following packages (and we have added a few convenience imports) |
HTML · PDF · Notebook |
|
ECON526: Midterm 2 Practice Ungraded practice problems for Midterm 2. |
HTML · PDF · Notebook |
No matching items