SGN-41007 Pattern Recognition and Machine Learning, 5 cr
Additional information
For more details, see last year slides and videos at http://www.cs.tut.fi/kurssit/SGN-41007/
The course substitutes earlier course SGN-41006 Signal Interpretation Methods. Only one of SGN-41006 and SGN-41007 can be accepted.
Kurssi korvaa aiemman kurssin SGN-41006 Signal Interpretation Methods. Vain toinen kursseista SGN-41006 ja SGN-41007 voidaan suorittaa.
Suitable for postgraduate studies.
Person responsible
Heikki Huttunen
Lessons
Implementation | Period | Person responsible | Requirements |
SGN-41007 2019-01 | 2 |
Heikki Huttunen |
Accepted exercises, course assignment and final exam. |
Learning Outcomes
Students understand principles of selected statistical, pattern recognition and machine learning approaches in signal processing related problems. Student can apply the methods to real problems using modern Python tools such as Scikit-Learn and Keras. For more details, see last year slides and videos at http://www.cs.tut.fi/kurssit/SGN-41007/
Content
Content | Core content | Complementary knowledge | Specialist knowledge |
1. | Statistical Signal Processing: Estimation theory; Maximum likelihood; Estimation of signal parameters (e.g., phase, amplitude and frequency). | ||
2. | Detection theory; Receiver Operating Characteristics; Neyman-Pearson decision rule and relation to machine learning. | ||
3. | Linear models: regression and classification, support vector machines, logistic regression, regularization. | ||
4. | Modern tools: Random forests, Bagging, Boosting, Stacking, Deep Learning | ||
5. | Performance evaluation, cross-validation, bootstrapping | ||
6. | Implementations in Python: 1) Scikit-learn, 2) Keras |
Instructions for students on how to achieve the learning outcomes
Accepted exercises and assignment. Final exam.
Assessment scale:
Numerical evaluation scale (0-5)
Partial passing:
Study material
Type | Name | Author | ISBN | URL | Additional information | Examination material |
Lecture slides | Pattern Recognition and Machine Learning | Heikki Huttunen | Yes |
Prerequisites
Course | Mandatory/Advisable | Description |
SGN-13006 Introduction to Pattern Recognition and Machine Learning | Mandatory | 1 |
1 . Either SGN-13000 or SGN-13006 is a prerequisite.
Additional information about prerequisites
The students are assumed to have the basic skills in probability, matrices and programming. Also the fundamentals of ML theory (SGN-13000 or SGN-13006) is strongly recommended.
Correspondence of content
Course | Corresponds course | Description |
SGN-41007 Pattern Recognition and Machine Learning, 5 cr | SGN-41006 Signal Interpretation Methods, 4 cr |