|
|
|||||||||||||||||
Course Catalog 2013-2014
SGN-90006 Signal Processing Doctoral Seminar, 2-8 cr |
Additional information
Suitable for postgraduate studies
Person responsible
Ioan Tabus, Ari Visa, Karen Eguiazarian, Ulla Ruotsalainen
Lessons
| Study type | P1 | P2 | P3 | P4 | Summer | Implementations | Lecture times and places |
|
|
|
|
|
|
|
|
|
Learning Outcomes
- Is familiar with a specialized area of signal processing where there is active research worldwide - Is able to critically evaluate the content of a specialized textbook in signal processing area - Acquires practice skills in preparing a presentation directed to a specialist audience - Acquires practice skills in defending his views when subject to criticism from the audience - Acquires practice skills in asking questions during the presentations of scientific topics
Content
| Content | Core content | Complementary knowledge | Specialist knowledge |
| 1. | Seminar topics: signal processing and related topics, with emphasis on algorithms. |
Prerequisite relations (Requires logging in to POP)
Correspondence of content
| Course | Corresponds course | Description |
|
|
|
More precise information per implementation
| Implementation | Description | Methods of instruction | Implementation |
| SGN-90006 Signal processing doctoral seminar, 3cr 3rd period, Wednesdays 10-12 at TB220 (TB215 on 15.1.2014 and TB214 on 5.2.2014 exceptionally) First meeting 8.1.2014 Instructors: Olli Yli-Harja Instructor reception on wednesdays 14-16 at TE306 Requirements: One student presentation and participation in lectures and presentations Registration: By email to yliharja@cs.tut.fi or at the first lecture Challenges in computational modeling and machine learning The focus of the seminar is to explore limitations of machine learning, basic assumptions on which mathematical modeling is based on, and principled limitations of computational approaches in science. The seminar is composed of a series of invited lectures ranging from statistical inference, machine learning to the pragmatistic approach to the philosophy of science, and student presentations on selected topics. Individual guidance sessions with the instructor are available for students to help the preparation of their presentations. The seminar concentrates on principled properties and limitations of computational approaches instead of details of algorithmic methods. The aim is to reveal useful viewpoints that are applicable in practical engineering work, e.g. related to "big data". We study success cases in machine learning as well as approaches which run into unavoidable difficulties. |