Probability and Statistics 2 (NMAI073)

Instructor: Robert Šámal

Coordinates
Lecture: Mon 17:20 in S3.
Tutorial: Tue 12:20 and 14:00 (Czech version for part of the semester) in S10.
Homeworks: in the Postal Owl
Syllabus
See info in SIS.
Last year version
To get a rough idea about the class, you may see the last year's version with all materials. (There will be some changes though.)
Exam
Will be organized similarly as in Prob&Stat1: written test from both computing problems and theoretical understanding. It will be graded 1-5. After that, you will have a possibility to take oral exam to improve by one grade.
If you solve more homework problems that the required 50 %, you will be getting extra points for the exam test (up to 10 extra points if you get full score for homework problems).
Credit (Zapocet) is not required for the first attempt to take the exam.
You may bring a "formula sheet" of your own notes on a single sheet of A4 paper (doublesided).
exam from Jan 19
Literature
See info in SIS.
Lecture notes (version Jan 29) (final version for this term)

Personal notes of Jonáš Havelka from last year -- in Czech only, no guarantee from me, nor from the author. It looks helpful though, for a review etc.
Log of lectures
Lecture 1 -- Oct 2, 2023
Definition of a Markov chain, examples. Multistep transition probabilities (Kolmogorov-Chapman recursion formula). Accesible states.
video
Lecture 2 -- Oct 9, 2023
Probability of particular sequence of states. General form of indepednece of future on history. Communicating states, equivalence relation. Reduction to smaller Markov chain. Classification of states: recurrent, transient. Theorem about convergence to stationary distribution. (just stated)
video
Lecture 3 -- Oct 16, 2023
Discussion of the theorem about convergence. Probability of absorption, time to absorption. Application: probabilistic algorithm for 2-SAT (to be finished next time).
video
Lecture 4 -- Oct 23, 2023
Randomized algo for 2-SAT and 3-SAT. What is probability? Intro to Baysian statistics: basic principles of Bayesian inference.
video
youtube video on Bayesian principles
Lecture 5 -- Oct 30, 2023
MAP and LMS methods to get point estimates. Baysian statistics -- example computations using Bayes formula.
video
Lecture 6 -- Nov 6, 2023
Bayesian statistics -- Conjugate gradients. Comments on MCMC. Conditional expectation (as a random variable) -- Law of iterated expectation
video
Lecture 7 -- Nov 13, 2023
Conditional expectation (as a random variable) -- proof of Law of iterated expectation, cond. exp. as an estimator, Eve's rule for variance.
Pdf, joint and conditional pdf revisited.
video
Lecture 8 -- Nov 20, 2023
Balls&bins: birthday paradox, upper bound for maxload, applications (hashing and bucketsort).
video
calculator for the birthday paradox
Lecture 9 -- Nov 27, 2023
Balls&bins: Poisson approximation. Simulation shown in the class as python notebook and as pdf.
Bernoulli processes: very brief introduction.
video
Lecture 10 -- Dec 4, 2023
Bernoulli and Poisson process.
video
Lecture 11 -- Dec 11, 2023
Permutation test. Nonparametric statistics: Permutation test, Sign test, Wilcoxon's signed rank test.
video
illustration of the problem
The animation I showed comes from wiki page on Permutation tests.
A nice visual explanation
Lecture 12 -- Dec 18, 2023
Permutation test -- results for the motivational example Simpson's paradox Shannon Source Coding theorem & entropy.
video
Lecture 13 -- Jan 8, 2024
Central limit theorem -- proof using moment generating functions.
Proof of Chernoff bound and its applications.
video