courses:m33sad:program [CourseWare Wiki]

The lectures are given in English to all students.

L	Date	Lecturer	Contents	Material
1	3.10.	JK	Intro to unsupervised learning, cluster analysis - formalization	sad_intro.pdf
2	10.10.	JK	K-means algorithm, EM GMM, hierarchical clustering	clustering.pdf
3	17.10.	JK	Advanced clustering methods – spectral clustering	clustering_pokr.pdf
4	24.10.	JK	Other clustering approaches – co-clustering, conceptual and semisupervised clustering	see the previous lecture
5	31.10.	JK	Frequent itemsets, association rules, the Apriori algorithm	apriori_eng.pdf
6	7.11.	JK	Frequent sequences and subgraphs	seq_graphs.pdf
7	14.11.	JK	Dimensionality reduction	dimreduction.pdf
8	21.11.	FŽ	Intro to supervised learning and computational learning theory	colt.pdf
9	28.11.	FŽ	Continuing Lecture 1
10	5.12.	FŽ	Learning in propositional logic	pac-logic.pdf
11	12.12.	FŽ	Continuing Lecture 3
12	19.12.	FŽ	Infinite Concept Spaces (only the 1st file required for exam)	infspaces.pdf, infspaces2.pdf
13	2.1.	FŽ	Empirical testing of hypotheses	empirical.pdf
14	9.1.	FŽ	Learning in predicate logic	predicate.pdf

Additional reading: see http://www.cs.princeton.edu/~mona/MachineLearning_lecture_notes.html (especially lectures 4-8) for a detailed treatment of PAC learnability.

T	Date	Deadline	Contents	Materials
1	10.10.		Introduction, program, requirements, SW; entrance test (prerequisite 33RPZ)	PDF
2	17.10.		Missing Values and Outliers; Removing Outliers using k-means Algorithm	PDF ZIP
3	24.10.		EM Algorithm and Semi-Supervised Learning	PDF ZIP
4	31.10.		Spectral Clustering	PDF ZIP
5	7.11.		Frequent Itemsets, Association Rules	PDF ZIP
6	14.11.		Dimensionality reduction	PDF ZIP
7	21.11.		Preparation for Test	vzorovy_test.pdf
8	28.11.		Test
9	5.12.	12.12.	Underfitting and Overfitting, Learning Curves	PDF ZIP
10	12.12.	19.12.	Upper Bounds for Classification Errors	PDF ZIP
11	19.12.	2.1.	Learning k-term DNF using k-CNF	PDF ZIP
12	2.1.	9.1.	Parameter Selection using Cross-Validation	PDF ZIP
13	9.1.		Closing all submissions & Credit