Statistics for Particle Physicists LPC 2021 Recordings - HATS@LPC and LPC Academic Lectures
"Statistics in Particle Physics” is a remote-only class taught by Prof. Harrison Prosper of the Florida State University for the LPC. The class meets over Zoom with CMS students on Mondays and Wednesdays 3:00-4:30 PM (central time), from October 4th, 2021 to December 1st, 2021. The organizers and Prof. Harrison Prosper have generously agreed to have the video recordings and class materials be made public after the lectures. They are posted below.
Those with CMS CERN accounts may register and attend the course at the Indico Agenda.
Video recordings
Note: these links include associated lecture files, homework, and homework solutions - see Additional files
- Lecture 1, October 4, 2021 - Probability Part 1
- includes handout01 and assignment1
- October 11, 2021 added solution1
- Lecture 2, October 6, 2021 - Probability Part 2
- includes handout02 and birthday_problem.ipynb
- October 11, 2021 added ratio_of_two_gaussian_variates.ipynb, likelihood_fit.ipynb, and least_squares_fit.ipynb
- Lecture 3, October 11, 2021 - Overview of Statistical Inference
- includes handout03 and assignment2
- October 18, 2021 added solution2
- October 19, 2021 added product_of_two_gaussian_variates.ipynb
- November 10, 2021, fixed link which went to wrong video
- Lecture 4, October 13, 2021 - Overview of Machine Learning
- Lecture 5, October 18, 2021 - Relative Frequency
- includes handout05 and assignment3
- October 25, 2021 added solution3
- Lecture 6, October 20, 2021 - Frequentist Inference - Part 1
- includes handout06
- October 25, 2021 added likelihood_fit_iminuit.ipynb, likelihood_data.txt, and FeldmanCousinsIntervals.ipynb
- Lecture 7, October 25, 2021 - Frequentist Inference - Part 2
- includes handout07 and assignment4
- November 10, 2021 - added solution4
- Lecture 8, October 27, 2021 - Frequentist Inference - Part 3
- includes handout08 (updated Oct. 29, 2021)
- Lecture 9, November 1, 2021 - Frequentist Inference - Part 4
- includes handout09 and assignment5
- November 10, 2021 added solution5
- Lecture 10, November 3, 2021 - Bayesian Inference - Part 1
- Lecture 11, November 8, 2021 - Bayesian Inference - Part 2
- includes handout11, assignment6
- November 15, 2021, updated handout11
- November 16, 2021, added solution6
- Lecture 12, November 10, 2021 - Bayesian Inference - Part 3
- Lecture 13, November 15, 2021 - Bayesian Inference - Part 4
- includes handout13
- November 17, 2021, added assignment7 and project.tar.gz
- December 6, 2021, added solution7.pdf and solution7.ipynb
- Lecture 14, November 17, 2021 - Machine Learning - Part 1
- includes handout14_v1 and boosting, updated morning November 18, 2021
- Lecture 15, November 29, 2021 - Introduction to Machine Learning - Part 2
- includes handout15
- Added cnn.tar.gz "which unpacks into three files, a jupyter notebook and two compressed MNIST data files. The notebook is an inplementation of the CNN we discussed in class today." (added on Nov. 20, 2021)
- Lecture 16, December 1, 2021 - Machine Learning - Part 3
Syllabus
This course introduces statistical concepts and methods to graduate students in particle physics. We assume no familiarity with statistics and develop the most important concepts and methods in detail, but err always on the side of clarity. In order to avoid complications that merely obscure, we shall, for the most part, use simple examples. If, by the end of the course, you have diligently worked through the examples on your own—or in good faith collaboration with your fellow students, you should be able to perform your own statistical calculations essentially from scratch. Of course, we would not expect you to write your own version of Minuit (a program created by CERN scientist Fred James for minimizing functions)! But, we would expect you to be able to replicate results of the standard statistics packages using your own programs, at least for a few of the standard problems in particle physics. If you are able to do this by the end of this course, we shall have succeeded. The only way to master something is to do it yourself. Consequently, homework is a vital component of this course. Your course grade is determined solely by your graded homework.
Prerequisites: College-level algebra, calculus, a modicum of residue theory, and basic programming skills. (If you do not know how to code, learning Python is easy and highly recommended. Also, knowing a bit of the CERN package ROOT would be extremely handy.)
Questions or suggestions can be emailed to the LPC coordinators
Webpage maintained by Marguerite Tonjes.