- Instructor: Jordan Bryan (jbryan@virginia.edu)
- Lecture: Tues, Thurs 9:30 - 10:45 am, Data Science Building Room 246
- Office hours: Mon 2:00 - 3:00 pm, Data Science Building Room 347
- Teaching Assistant: Elizabeth Miller (zrc3hc@virginia.edu)
- Office hours: Fri 9:30 - 10:30 am, Data Science Building Room 300
- Canvas site
-
Computer Age Statistical Inference: Algorithms, Evidence, and Data Science (CASI)
-
Statistical Inference (Casella & Berger)
-
Course code snippets
- Foundations of inference
- Sampling distributions
- Point estimation, interval estimation, and hypothesis testing
- Optimality criteria
- Likelihood methods
- Parametric models and exponential families
- Information
- Maximum likelihood
- Asymptotic evaluations
- Consistency
- Delta method
- Asymptotic properties of MLEs
- Bootstrapping and resampling methods
- The jackknife estimate of standard error
- Non-parametric bootstrap
- Parametric bootstrap
- Prediction
- Estimates of predictive accuracy
- Shrinkage and ridge regression
- Contemporary ideas in prediction (if time allows)
Jan 13 2026: Course overview + the terminology of inference
Jan 15 2026: Sample sizes tending toward infinity
- HW1 assigned (due 01/22/2026 at 9:30 am)
Jan 20 2026: Beyond the CLT, analyzing estimators
[//]: # Consistency Slutsky's theorem, Taylor approximation, Delta Method
Jan 22 2026: Bias and variance
- HW1 explication: Stephanie and Ziqian
- HW2 assigned (due 01/29/2026 at 9:30 am)
Jan 27 2026: Bounds for optimal estimation, information
Jan 29 2026: Data reduction, sufficiency
- HW2 explication: Andres and Cynthia
- HW3 assigned (due 02/05/2026 at 9:30 am)
[//]: # Feb 03 2026: Interval estimation
Feb 05 2026: TBD
- HW3 explication: Habiba and Stephanie
- HW4 assigned (due 02/12/2026 at 9:30 am)
[//]: # Feb 10 2026: Exponential families
Feb 12 2026: TBD
- HW4 explication: Ziqian and Andres
- HW5 assigned (due 02/19/2026 at 9:30 am)
[//]: # Feb 17 2026: Asymptotic properties of MLEs
Feb 19 2026: TBD
- HW5 explication: Cynthia and Habiba
- HW6 assigned (due 03/12/2026 at 9:30 am)
[//]: # Feb 24 2026: Midterm review
Feb 26 2026: Midterm exam
Mar 03-05 2026: Spring recess (no class)
[//]: # Mar 10 2026: TBD
Mar 12 2026: TBD
- HW6 explication: Stephanie and Andres
- HW7 assigned (due 03/19/2026 at 9:30 am)
[//]: # Mar 17 2026: TBD
Mar 19 2026: TBD
- HW7 explication: Habiba and Cynthia
- HW8 assigned (due 03/26/2026 at 9:30 am)
[//]: # Mar 24 2026: TBD
Mar 26 2026: TBD
- HW8 explication: Ziqian and Stephanie
- HW9 assigned (due 04/02/2026 at 9:30 am)
[//]: # Mar 31 2026: TBD
Apr 02 2026: TBD
- HW9 explication: Andres and Habiba
- HW10 assigned (due 04/09/2026 at 9:30 am)
[//]: # Apr 07 2026: TBD
Apr 09 2026: TBD
- HW10 explication: Cynthia and Ziqian
- HW11 assigned (due 04/16/2026 at 9:30 am)
[//]: # Apr 14 2026: TBD
[//]: # Apr 16 2026: TBD HW11 assigned
[//]: # Apr 21 2026: TBD
[//]: # Apr 23 2026: TBD
Apr 28 2026: Final review
May 04 2026: Final exam (2:00 - 5:00 pm, location TBD)
Final grades will be computed using the following weighting:
- Attendance (5%)
- Homework (30%)
- Homework explications (15%)
- Midterm exam (20%)
- Final exam (30%)
Grading scale:
- 93-100 A
- 90-92 A-
- 87-89 B+
- 83-86 B
- 80-82 B-
- 77-79 C+
- 73-76 C
- 70-72 C-
- <70 F
Note that a B- is the lowest satisfactory grade for graduate credit.
Submitting Homework
Homework will be accepted through the Assignments page on Canvas. Submissions will be in PDF format. You may hand-write and scan problem solutions, or you may use a typesetting software like LaTeX, Markdown, etc. Some homework assignments will involve using code to produce graphical or numerical outputs and will require the use of software. Please compile all materials in a single PDF for submission and make sure that whatever you have written can be clearly read by the grader.
Grades for (on-time) homework will be made visible to students no later than one week after the assignment due date. Grades for late work (see below) will become available as time permits.
Late Work Policy
The expectation in this course is that all assignments will be submitted on time. Submitting your work on time respects the efforts of your instructor and teaching assistant, and it ensures that you are prepared to learn subsequent material.
Assignments turned in after the due date incur a 10% penalty per late day. For example, an assignment due at 9:30 am on Tuesday that is submitted to Canvas at 3:00 pm on Thursday will incur a 30% penalty. If the assignment would have received a 95% had it been returned on time, then the late grade is 65%. Note that weekend days count towards the late penalty.
I will not accept work that is late by more than one week past its due date.
To provide flexibility for weeks in which life circumstances do not permit the completion of your coursework, your lowest homework grade will be dropped.
Class Attendance
Attendance in this class is mandatory. If you need to miss a class for any reason, please email me in advance. You are responsible for keeping up with the lecture material, but I am happy to work with you during office hours or by appointment to brush up on things you may have missed.
Extenuating Circumstances
Students are expected to communicate with me as soon as possible regarding extenuating circumstances and how their participation in the course, including attendance and assignment submissions, may be affected by them.
Academic Integrity
I encourage collaboration among students to complete homework assignments. The purpose of collaborating is to help yourself and your classmates learn the material more effectively.
Do not cheat. Cheating circumventes the learning process and deprives you of the chance to gain expertise in your discipline. It also puts you in a position to fail the in-class exams, for which you will not be able to use resources outside of your own problem solving ability
I ask that you:
- Do not copy text or code from classmates, the internet, or AI systems. Write your own solutions and understand them.
- Do not send text or code to classmates or post your solutions in a place where everyone can access them. If you are collaborating with others, you are working together to arrive at a solution.
- Try to solve each problem before resorting to outside help. Even if you do not ultimately arrive at a solution completely on your own, starting the solution process by yourself is crucial to learning.
If an action you are considering is not covered by one of these specific asks, please use your own sense of right and wrong to determine whether it constitutes cheating.
Definition and goal
An explication is a detailed explanation. The goal of homework explications is to give students the opportunity to explain, in detail, their reasoning as they work towards the solution of a homework problem. Explications are intended to encourage real-time logical thinking and argument based upon the principles learned in class and in readings.
How will explications be implemented?
For each homework that is assigned, two students will be asked to give an explication. Explications will take place during the class period coinciding with the assignment due date (typically this will be a Thursday). Specifically,
-
At the start of class time, the instructor will choose a homework problem from the week's assignment and ask one of the students to present their solution to the problem to the class on the whiteboard (if an element of the solution involves a simulation, the student may use a laptop as well).
-
A Q&A and discussion will follow, in which members of the class and/or the instructor may ask clarifying questions of the presenting student. Subjects of relevance to lecture material will be elucidated.
-
Steps 1. and 2. will be repeated for the second student, but a different problem will be chosen. The two explications combined should take around 25-30 minutes of class time.
How will explications be graded?
Your grade will be determined by the instructor based on a combination of
- The quality of your solution.
- The clarity of your reasoning.
- Your ability to justify your solution choices based on things we have learned in class and in readings.
Note that the correctness of your solution is not a criterion. Yes, correct solutions may tend to be higher quality, but an incorrect solution could still be high quality if the steps taken along the way are coherent and generally sensible.
Additional notes for explicators
-
You may use your notes (i.e. your homework solutions) as you write and explain your solution. However, please do not treat this as an exercise in rote memorization. You need to understand your solutions in order to answer questions about them.
-
You should be prepared to explain your solution to any of the homework problems from the current week's assignment. You will not know in advance which solution you will be asked to explain.