This course is about methodologies and algorithms allowing computer programs to learn from data (examples) and by this become better in solving some given task. The course covers the main principles of machine learning, concentrating on supervised learning algorithms. Acquaintance will be made with the main types of models, such as decision trees, linear models, distance-based models, probabilistic models, ensemble models and neural networks. Practical assignments will be given to build a deeper understanding of the material.
- The course starts on September 10, 2018 (no lecture and no practice sessions during the week September 3-7)
- Lectures: Monday 10:15 - 12:00, J.Liivi 2, room 111
- Practice Sessions:
- Group 1: Monday 12:15 - 14:00 (J.Liivi 2, room 402, Mikhail)
- Group 2: Monday 12:15 - 14:00 (J.Liivi 2, room 405, Meelis)
- Group 3: Monday 14:15 - 16:00 (J.Liivi 2, room 122, Mikhail)
- Group 4: Monday 14:15 - 16:00 (J.Liivi 2, room 405, Meelis)
- Group 5: Friday 14:15 - 16:00 (J.Liivi 2, room 402, Mikk)
- Lecturer: Meelis Kull (firstname.lastname@example.org)
- Teaching Assistants: Mikhail Papkov (email@example.com) and Mikk Puustusmaa (firstname.lastname@example.org)
- Test 1 - Oct 15, 10:15-11:45
- Test 1 resit - Nov 5, 18:15-19:45
- Test 2 - Nov 19, 10:15-11:45
- Test 2 resit - Dec 10, 18:15-19:45
- Test 3 (limited number of students with pre-registration) - Dec 20, 16:15-18:00
- Test 3 - Jan 7, 9:15-11:00
- Test 3 resit - Jan 21, 16:15-18:00
Locations of tests will be announced on Piazza (all tests will be in J.Liivi 2).
- Forum and discussions will be held in Piazza. If you didn't get an invitation email by September 11, please let us know!
Grading and requirements:
The grade is calculated from the total number of points (max 100). The points can be earned as follows:
- Homeworks (36 points): there will be 6 homeworks, each worth 6 points;
- Tests (64 points): there will be 3 tests worth of 20, 20 and 24 points.
In order to pass the course the student must get at least 50% from homeworks (threshold 18 points) and 50% from each of the tests (thresholds 10, 10 and 12 points, respectively).
There will be no exam. The final grade is calculated as explained above.
Homework and test results are visible here
The course will be partly based on the book
"Machine learning: The Art and Science of Algorithms that Make Sense of Data" by Peter Flach (2012)
Course programming language:
Homeworks are required to be solved using Python version 3. Practice sessions, examples and support is given only for that language.