2-INF-150: Machine learning (Strojové učenie)
Winter 2018
Prednášky a poznámky


Contacts | Basic information | Homework assigments | Lectures and handouts | Older versions of this course


Here you can find preliminary schedule for semester. This schedule would be updated after finishing particular week of lectures.

Ďalšie zdroje informácií:

We provide parts of literature and external links for each leacture. Prezentation in these materials might be different than presentation during the lecture. There materials should be treated as supplementary.

Week 24.-28.9.2018
Admnistration. Introduction. Supervised learning. Linear regression. Basic math (gradients, partial derivatives, matrices).
Literature: GBC:2.1-2.4; GBC:5.1
Slajdy a poznámky:Ďalšie materiály:
slides 1:PDF, 135 Kb ]
slides 1b:PDF, 225 Kb ]
slides 2:PDF, 158 Kb ]
slides 2(english):PDF, 157 Kb ]
slides 3a:PDF, 252 Kb ]
slides 3a(english):PDF, 250 Kb ]
stanford (chapters 1, 2, 4):linka ]

Week 1.-5.10.2018
Theory of ML, overfitting, bias, variance.
Tutorials: numpy
Literature: GBC:2.1-2.4; GBC:5.1
Slajdy a poznámky:Ďalšie materiály:
Algebra 101:PDF, 42 Kb ]
Tutorials 1: numpy:linka ]
stanford (chapter 4):linka ]
stanford (chapter 1):linka ]

Week 8.-12.10.2018
Theory of ML, bias variance decomposition. Regularization. Classification. Logistic regression, softmax (maximum entropy) classifier. Probabilistic interpretion of regression.
Slajdy a poznámky:Ďalšie materiály:
stanford (chapter 3, 5 and 9.3):linka ]
stanford (chapter 1):linka ]

Week 15.-19.10.2018
Neural networks.
Tutorials 2: regression
Slajdy a poznámky:Ďalšie materiály:
Tutorials 2: regression:linka ]

Week 22.-26.10.2018
Support vector machines.
Tutorials 3: neural networks
Slajdy a poznámky:Ďalšie materiály:
Tutorials 3: neural networks:linka ]
Neural nets slides:PDF, 11237 Kb ]

Week 29.-2.11.2018
Free week.

Week 5.-9.11.2018
Support vector machines (continued). Decision trees
Slajdy a poznámky:Ďalšie materiály:
SVM Stanford:linka ]
Decision Trees:linka ]
Decision Trees 2:linka ]

Week 12.-16.11.2018
Bagging and boosting.
Tutorials 4: trees and SVMs
Slajdy a poznámky:Ďalšie materiály:
Tutorials 4: trees and SVMs:linka ]
Gradient boosting:linka ]

Week 20.-24.11.2017
Theory of ML (PAC learning). Estimating needed number of training samples and expected test error for finite hyphothesis set. Estimation for infinite set of hyphotheses. VC dimension. PAC learning and SVM.
Slajdy a poznámky:Ďalšie materiály:
PAC - finite hyphotheses:PDF, 243 Kb ]
axis aligned rectangles / Andres Munoz, NY Univ:PDF, 97 Kb ]
VC dimension - definition and examples / Yishai Mansour, U Tel Aviv:PDF, 109 Kb ]
VC dimension - PAC estimates / Yishai Mansour, U Tel Aviv:PDF, 173 Kb ]
PAC estimates for SVM:PDF, 295 Kb ]


Maintained by 2-INF-150 personnel