Course syllabus

SF2957 HT24 Statistical Machine Learning

Lecturer: Viktor Nilsson. Email: vikn@kth.se

Teaching assistant: Hanqing Xiang. Email: hanqingx@kth.se

First Lecture: The first lecture will take place on Tuesday, October 29, 10:00-12:00 in H1.

Course Layout: The course will consist of two (sometimes three) lectures and one exercise class per week. Toward the end of the course, there will be two scheduled activities for presenting the assigned projects.

Course literature: Lecture notes and extracts from various books

• Casella, G. and Berger, R.L., Statistical inference, 2nd Ed., Duxbury Thomson Publishing, 2002.
• Goodfellow, I., Bengio, Y., and Courville, A., Deep learning, MIT Press, 2016. Available electronically at: https://www.deeplearningbook.org Links to an external site.
• MacKay, D.,  Information theory, inference, and learning algorithms, Cambridge University Press, 2003. Available as PDF at: http://www.inference.org.uk/itprnn/book.pdf Links to an external site.
• Rasmussen, C.E. and Williams, K.I., Gaussian processes for machine learning, MIT Press, 2006. Available as PDF at: http://www.gaussianprocess.org/gpml/chapters/RW.pdf Links to an external site.
• Sutton, R.S. and Barto, A.G., Reinforcement learning: an introduction, 2nd Ed., MIT Press, 2018. Available as PDF at:  http://incompleteideas.net/book/RLbook2020.pdf Links to an external site.
• Kushner, H. J. and Yin, G. G., Stochastic approximation and recursive algorithms and applications, 2nd Ed., Springer, 2003. 
• Schervish, M. J., Theory of statistics, Springer, 1995

Sparsely, I will also refer to:

• Duchi, J.C., Introductory lectures on stochastic optimization, The mathematics of data 25, 2018. Available as PDF at: https://stanford.edu/~jduchi/PCMIConvex/Duchi16.pdf Links to an external site.
  
• Bishop, C.M., Pattern recognition and machine learning, Springer, 2006. Available as PDF at: https://www.microsoft.com/en-us/research/uploads/prod/2006/01/Bishop-Pattern-Recognition-and-Machine-Learning-2006.pdf Links to an external site.
  
• Gelman, A. et al., Bayesian data analysis, 3rd Ed., Chapman & Hall, 2013. Available as PDF at: http://www.stat.columbia.edu/~gelman/book/BDA3.pdf Links to an external site.

Recommended additional reading can be found here (for more context, insights, viewpoints, explanations, etc., at your own appetite).

Learning outcomes: This course presents an overview of advanced methods of statistical machine learning. The course addresses both theoretical and practical aspects of statistical machine learning. Computer-aided projects with a variety of datasets form an essential learning activity. To complete the course, the student must be able to:

• formulate and apply statistical decision theory
• formulate and apply advanced methods in statistical machine learning
• design and implement advanced methods in statistical machine learning for applications

Examination: The examination consists of projects (3.5 credits) and a final exam (4 credits).  For a passing grade, it is required to obtain a pass on the projects as well as the final exam. Registration for the final exam is compulsory. You can register at MyPages. Grades are given in the range A-F and Fx. The grade Fx gives you the right to a complementary examination to reach the grade E. The criteria for Fx is F and an isolated part of the course can be identified where you have shown a particular lack of knowledge and the examination after a complementary examination on this part can give the grade E. 

Grading: The requirements for higher grades are as follows. 

Projects: The projects examine the third learning outcome. Project information will be available on the course's Canvas page. Participation in a project group and writing a report is mandatory to pass the course. 

•You must work in groups of at most four people.

•You must work actively on the project together with the other group members and be prepared to account for your own and other group members’ contribution.

•You must hand in the report by uploading it on Canvas before the deadline.

•You must, together with your project group, be prepared to make a presentation of your project or part of your project.

Homework: There will be one hand-in assignment on the statistical inference part of the course. The homework can give a maximum of 4 bonus points for the first part of the exam.

Final Exam: The final exam takes place on Jan 8, 2025, 08:00-13:00. Registration for the final exam is compulsory. You can register at MyPages.

The final exam consists of three parts. Part I consists of two problems and examines the first learning outcome. Part II consists of two problems and examines the second learning outcome. Part III consists of one problem and examines higher-level learning outcomes. 

Support for students with disabilities: Information about support for students with disabilities (Funka) can be found here. 08-790 6911Länk