ECON 570: Big Data Econometrics
Hello, dear friend, you can consult us at any time if you have any questions, add WeChat: daixieit
ECON 570: Big Data Econometrics
Machine Learning and Causal Inference
Course Description
This course provides an introduction to the theory and practice of causal econometrics in modern settings of large-scale data. Major algorithms from machine learning will be introduced as tools for statistical pattern recognition as well as powerful aids to methods for identifying causal effects. While learnings from theory will be emphasized, the course will be focused on methodology and applications rather than rigorous proofs of theorems.
Learning Objectives
By the end of the course, students should be able to:
● Have a good sense of the computational complexity of different estimators and algorithms.
● Apply common machine learning algorithms on real data sets using a scientific computing software language.
● Develop and assess empirical strategies for identifying causal effects in both experimental and observational settings.
● Combine the tools of machine learning and causal inference to tackle empirical questions in big data settings.
Prerequisites
Necessary background for this course are calculus (at the level of MATH 226), linear algebra (at the level of MATH 225), and graduate-level econometrics (at the level of ECON 513). Formal training in causal inference and machine learning are not assumed, though some prior exposure is helpful.
Supplementary Materials
All required material will be covered by lecture notes. For students interested in more in-depth expositions, the following textbooks are recommended. Relevant chapters are listed in the course schedule.
1. Hastie, T., Tibshirani, R., and Friedman, J. (2009). The Elements of Statistical Learning. Springer. [HTF]
2. Bishop, C. M. (2007). Pattern Recognition and Machine Learning. Springer. [B]
3. Imbens, G. W., and Rubin, D. B. (2015). Causal Inference for Statistics, Social, and Biomedical Sciences: An Introduction. Cambridge University Press. [IR]
4. Angrist, J. D., and Pischke, J. (2009). Mostly Harmless Econometrics. Princeton University Press. [AP]
Course Grading
3 Problem Sets: 60%
Regular problem sets will be given to reinforce the concepts taught in class, as well as offer an opportunity for students to code and implement the algorithms covered. Students are encouraged to work in groups, but each person must turn in their own copy. Assessment will be based on whether the right approaches were used and whether the right solutions were obtained. Due dates for the assignments are September 24, October 15, and November 12.
1 Empirical Project: 40%
For the empirical project, students are expected to work in groups (maximum of four) and apply their learnings to a real data set to tackle an empirical problem that interests them. Each group must submit a short write-up (6 pages max) that summarizes the analysis, and computer code that reproduces the quantitative results. Assessment will be based on how appropriately the quantitative tools were applied. Due date for this project is December 3.
Software
Python will be the de facto programming language used in this course. Lecture notes will contain Python code snippets, TA sessions will include Python instructions, and any code in problem set solutions will be in Python. However, if students insist on doing their problem sets and project in another language like R, they may.
No prior exposure to Python is assumed. In fact the TA sessions and problem sets should guide students through the learning process. Prior exposure to some programming language is, however, extremely helpful.
Course Schedule
Week
|
Topics
|
Readings
|
1 |
Course overview; Computational complexity
|
|
2 |
Unsupervised learning: Dimensionality reduction
|
HTF 14
|
3 |
Unsupervised learning: Matrix factorization; Embeddings
|
HTF 14
|
4 |
Unsupervised learning: Cluster analysis
|
|
5 |
Supervised learning: Model selection; Regularized regression
|
HTF 7, 5
|
6 |
Supervised learning: Tree-based methods
|
HTF 9
|
7 |
Supervised learning: Bagging; Boosting
|
HTF 15, 10
|
8 |
Introduction to Deep Learning
|
|
9 |
Causal inference: Unconfoundedness; Propensity score methods
|
IR 12-17
|
10 |
Causal inference: Matching estimators
|
IR 18
|
11 |
Causal inference: Instrumental variables; Regression discontinuity
|
IR 23, 24; AP 6
|
12 |
ML econometrics: Flexible controls; heterogeneous treatment effects
|
|
13 |
Buffer
|
|
Statement on Academic Conduct and Support Systems
Academic Conduct:
Plagiarism – presenting someone else’s ideas as your own, either verbatim or recast in your own words – is a serious academic offense with serious consequences. Please familiarize yourself with the discussion of plagiarism in SCampus in Part B, Section 11, “Behavior Violating University Standards” policy.usc.edu/scampus-part-b. Other forms of academic dishonesty are equally unacceptable. See additional information in SCampus and university policies on scientific misconduct, http://policy.usc.edu/scientific-misconduct.
Support Systems:
Student Counseling Services (SCS) – (213) 740-7711 – 24/7 on call
Free and confidential mental health treatment for students, including short-term psychotherapy, group counseling, stress fitness workshops, and crisis intervention. engemannshc.usc.edu/counseling
National Suicide Prevention Lifeline – 1 (800) 273-8255
Provides free and confidential emotional support to people in suicidal crisis or emotional distress 24 hours a day, 7 days a week. www.suicidepreventionlifeline.org
Relationship and Sexual Violence Prevention Services (RSVP) – (213) 740-4900 – 24/7 on call
Free and confidential therapy services, workshops, and training for situations related to gender-based harm. engemannshc.usc.edu/rsvp
Sexual Assault Resource Center
For more information about how to get help or help a survivor, rights, reporting options, and additional resources, visit the website: sarc.usc.edu
Office of Equity and Diversity (OED)/Title IX Compliance – (213) 740-5086
Works with faculty, staff, visitors, applicants, and students around issues of protected class. equity.usc.edu
Bias Assessment Response and Support
Incidents of bias, hate crimes and microaggressions need to be reported allowing for appropriate investigation and response. studentaffairs.usc.edu/bias-assessment-response-support
The Office of Disability Services and Programs
Provides certification for students with disabilities and helps arrange relevant accommodations. dsp.usc.edu
Student Support and Advocacy – (213) 821-4710
Assists students and families in resolving complex issues adversely affecting their success as a student EX: personal, financial, and academic. studentaffairs.usc.edu/ssa
Diversity at USC
Information on events, programs and training, the Diversity Task Force (including representatives for each school), chronology, participation, and various resources for students. diversity.usc.edu
USC Emergency Information
Provides safety and other updates, including ways in which instruction will be continued if an officially declared emergency makes travel to campus infeasible. emergency.usc.edu
USC Department of Public Safety – UPC: (213) 740-4321 – HSC: (323) 442-1000 – 24-hour emergency or to report a crime.
Provides overall safety to USC community. dps.usc.edu
2021-09-30
Machine Learning and Causal Inference