T-61.181 Special course in Information Science I

INDEPENDENT COMPONENT ANALYSIS, AUTUMN 2001

TENTATIVE LECTURE SCHEDULE AND COURSE PLAN

Date	Topic	Lecturer
Lecture 1 17.9.	General information. Chapter 1: Introduction	Erkki Oja
	Chapter 2: Random vectors and independence, part 1	Juha Karhunen
Lecture 2 24.9.	Chapter 2: Random vectors and independence, part 2	Juha Karhunen
	Chapter 3: Gradients and optimization methods	Erkki Oja
Lecture 3 01.10.	Chapter 4: Estimation theory	Juha Karhunen
	Chapter 5: Information theory	Juha Karhunen
Lecture 4 08.10.	Chapter 6: Principal component anal. and whitening	Erkki Oja
	Chapter 7: What is independent component analysis?	Erkki Oja
Lecture 5 15.10.	Chapter 8: ICA by maximization of nongaussianity	Juha Karhunen
	Chapter 8: ICA by maximization of nongauss., part 2	Juha Karhunen
Lecture 6 25.10.	Chapter 9: ICA by maximum likelihood estimation	Erkki Oja
	Chapter 10: ICA by minimization of mutual inform.	Erkki Oja
Lecture 7 29.10.	Chapter 11: ICA by tensorial methods	Erkki Oja
	Chapter 12: ICA by nonlinear decorrelation and ...	Erkki Oja
Lecture 8 5.11.	Chapter 13: Practical considerations	Erkki Oja
	Chapter 14: Overview and comparison of basic ...	Erkki Oja
Lecture 9 12.11.	Chapter 15: Noisy ICA	Juha Karhunen
	Chapter 16: ICA with overcomplete bases	Juha Karhunen
Lecture 10 19.11.	Chapter 17: Nonlinear ICA	Juha Karhunen
	Chapter 18: Methods using time structure, part 1	Juha Karhunen
Lecture 11 26.11.	Chapter 18: Methods using time structure, part 2	Juha Karhunen
	Chapter 19: Convolutive mixtures and blind ...	Juha Karhunen
Lecture 12 3.12.	Chapter 21: Feature extraction by ICA	Erkki Oja
	Chapter 22: Brain imaging applications	Erkki Oja

Chapters 20, 23, and 24 of the textbook A. Hyvärinen, J. Karhunen, and E. Oja, "Independent Component Analysis", J. Wiley 2001, 481+xxii pages, will be skipped. Lectures will be held in the room Y313 in the main building of HUT.