Statistical learning theory 2018 2019 — различия между версиями

Материал из Wiki - Факультет компьютерных наук
Перейти к: навигация, поиск
Строка 8: Строка 8:
 
Deadline homework 1: October 2nd. Questions: see seminars [https://www.dropbox.com/s/jb9mriumhtdpn8m/03sem.pdf?dl=0 3] and [https://www.dropbox.com/s/l2d9f7u77smrx4u/04sem.pdf?dl=0 4].  
 
Deadline homework 1: October 2nd. Questions: see seminars [https://www.dropbox.com/s/jb9mriumhtdpn8m/03sem.pdf?dl=0 3] and [https://www.dropbox.com/s/l2d9f7u77smrx4u/04sem.pdf?dl=0 4].  
  
Deadline homework 2: October 27nd.
+
Deadline homework 2: October 27nd. Questions: see seminars 5-8 below.
  
Deadline homework 3: TBA.
+
Deadline homework 3: TBA.  
  
 
[https://www.dropbox.com/s/dy9yu1ro4k5miet/List%20of%20Students_Bruno.xlsx?dl=0  Marks]
 
[https://www.dropbox.com/s/dy9yu1ro4k5miet/List%20of%20Students_Bruno.xlsx?dl=0  Marks]
Строка 37: Строка 37:
 
| 15 Oct || Rademacher complexity and contraction lemma (=Talagrand's lemma), Mohri's book pages 33-41 and 78-79 || [https://www.dropbox.com/s/y2vr3mrwp66cuvz/07lect.pdf?dl=0 lecture7.pdf] || [https://www.dropbox.com/s/cuo0tmfv4k2egvh/07sem.pdf?dl=0 Problem list 7] ||
 
| 15 Oct || Rademacher complexity and contraction lemma (=Talagrand's lemma), Mohri's book pages 33-41 and 78-79 || [https://www.dropbox.com/s/y2vr3mrwp66cuvz/07lect.pdf?dl=0 lecture7.pdf] || [https://www.dropbox.com/s/cuo0tmfv4k2egvh/07sem.pdf?dl=0 Problem list 7] ||
 
|-
 
|-
| 21 Oct || Margin theory and risk bounds for boosting. || ||
+
| 21 Oct || Margin theory and risk bounds for boosting. || || ||
 
|}
 
|}
  

Версия 17:57, 21 октября 2018

General Information

The syllabus

Questions colloquium on 29 October. (Lectures 1-5, lectures 6-8 follow later.)

Deadline homework 1: October 2nd. Questions: see seminars 3 and 4.

Deadline homework 2: October 27nd. Questions: see seminars 5-8 below.

Deadline homework 3: TBA.

Marks

Intermediate exams: October 29th.

Course materials

Date Summary Lecture notes Problem list Solutions
3 Sept PAC-learning in the realizable setting definitions lecture1.pdf updated 23/09 Problem list 1
10 Sept VC-dimension and growth functions lecture2.pdf updated 23/09 Problem list 2
17 Sept Proof that finite VC-dimension implies PAC-learnability lecture3.pdf updated 23/09 Problem list 3
24 Sept Applications to decision trees and threshold neural networks. Agnostic PAC-learnability. lecture4.pdf Problem list 4
1 Oct Agnostic PAC-learnability is equivalent with finite VC-dimension, structural risk minimization lecture5.pdf 14/10 Problem list 5
9 Oct Boosting, Mohri's book pages 121-131. lecture6.pdf 21/10 Problem list 6
15 Oct Rademacher complexity and contraction lemma (=Talagrand's lemma), Mohri's book pages 33-41 and 78-79 lecture7.pdf Problem list 7
21 Oct Margin theory and risk bounds for boosting.

A gentle introduction to the materials of the first 3 lectures and an overview of probability theory, can be found in chapters 1-6 and 11-12 of the following book: Sanjeev Kulkarni and Gilbert Harman: An Elementary Introduction to Statistical Learning Theory, 2012.

Afterward, we hope to cover chapters 1-8 from the book: Foundations of machine learning, Mehryar Mohri, Afshin Rostamizadeh, and Ameet Talwalker, 2012. These books can be downloaded from http://gen.lib.rus.ec/ .


Office hours

Person Monday Tuesday Wednesday Thursday Friday
Bruno Bauwens 16:45–19:00 15:05–18:00 Room 620


Russian texts

The following links might help students who have trouble with English. A lecture on VC-dimensions was given by K. Vorontsov. A course on Statistical Learning Theory by Nikita Zhivotovsky is given at MIPT. Some short description about PAC learning on p136 in the book ``Наука и искусство построения алгоритмов, которые извлекают знания из данных, Петер Флах. On machinelearning.ru you can find brief and clear definitions.