MLA310 Matrix Methods for Data Analysis and Machine Learning
Credits (ECTS):10
Course responsible:Ulf Geir Indahl
Campus / Online:Taught campus Ås
Teaching language:Norsk
Course frequency:The subject is not taught in autumn 2023 (Normally, the subject is taught every autumn semester)
Nominal workload:250 hours
Teaching and exam period:The course is implemented and censored each fall semester. The subject is not taught in autumn 2023
About this course
Derivation and applications of advanced matrix methods for pattern recoginition, machine learning and data analysis: The subjects include clustering, projection- and matrix factorization methods, variable selection and regularization for regression- and classification problems. We will also cover efficient computations for model selection and -validation.
Learning outcome
The students will learn both the theoretical background and how to implement various methods for advanced analysis of research data.
- The teaching will be given as lectures, practical exercises and project work.
- Oral exam based on the syllabus and mandatory project exercises.
- The censor will evaluate the examination including questions about the mandatory project work.
- Mandatory project exercises throughout the semester. Rules for approval of mandatory activity will be presented at the start of the semester.
- The subject is not taught in autumn 2023
- 4 hours per week including exercises.
- Letter grades