Please note! Course description is confirmed for two academic years, which means that in general, e.g. Learning outcomes, assessment methods and key content stays unchanged. However, via course syllabus, it is possible to specify or change the course execution in each realization of the course, such as how the contact sessions are organized, assessment methods weighted or materials used.
Schedule: 11.01.2021 - 09.04.2021
Teacher in charge (valid 01.08.2020-31.07.2022): Paavo Alku, Rohit Babbar, Tom Bäckström, Tom Bäckström, Alexander Ilin, Alex Jung, Juho Kannala, Samuel Kaski, Mikko Kurimo, Jouko Lampinen, Harri Lähdesmäki, Pekka Marttinen, Jussi Rintanen, Juho Rousu, Arno Solin, Aki Vehtari
Teacher in charge (applies in this implementation): Paavo Alku
Contact information for the course (valid 09.12.2020-21.12.2112):
Seminar, January-April 2021
Special Course in Machine Learning and Data Science: Signal
processing and machine learning methods for speech-based biomarking of human
health, 5 credits
- Pre-requisites: In order to register
to the course, the student must have basic knowledge in speech
processing. This means that the student must have taken either ELEC-E5500
Speech Processing or a similar course in another university. In the latter
case, the student must contact the teacher before registering to the course.
- The course is held as a seminar (in
zoom) where the teacher first gives an introduction lecture on the topic.
- After the introduction lecture, the
course continues in weekly seminars where students present selected articles on
the topic. The articles will be selected by the teacher and assigned to each
student. Each student gives 1-2 presentations. The student presentation includes:
(1) preparation of slides (about 10-20 pages) on the selected article, (2)
presenting the slides to the audience and (3) preparation of 2-3 simple
assignments on the presented article. (4) The student returns the model solution/answer
to the teacher.
- In order to pass the course, the
student must (a) give his/her own presentation, (b) must have answer correctly
to 80% of the assignments given by the other presenters and (c) must be present
in at least 80% of the presentations by other students.
- In order to get a
deeper understanding of the topic using real speech data, the students who have
passed the course are encouraged to continue by the following means: (1) MSc
(1.1): You have a possibility to take SPA-EV, a course with varying
contents (1-10 credits), see:
with the teacher and your own professor about possibilities to start a MSc
thesis on the topic,
(2) PhD students: select a topic of your own interest (which
is related to speech-based biomarking of human health) and continue studying
the topic by taking course ELEC029Z-LZ (1-10 credits), see https://www.aalto.fi/en/services/course-codes-at-school-of-electrical-engineering
CEFR level (applies in this implementation):
Language of instruction and studies (valid 01.08.2020-31.07.2022):
Teaching language: English
Languages of study attainment: English
CONTENT, ASSESSMENT AND WORKLOAD
The contents and implementation of the course vary, it can be lectured or arranged in seminar form.
Assessment Methods and Criteria
To be specified at the start of the course.
Usually some new book or collection of articles.
Substitutes for Courses
Substitutes course CS-E4070 Special Course in Machine Learning and Data Science
Sufficient courses on machine learning, data science and artificial intelligence
- Teacher: Paavo Alku