STAT729 MODERN DATA ANALYSIS: FROM HIDDEN MARKOV MODELS TO STATISTICAL LEARNING
| Course Code: |
2460729 |
| METU Credit (Theoretical-Laboratory hours/week): |
3(3-0) |
| ECTS Credit: |
8.0 |
| Department: |
Statistics |
| Language of Instruction: |
English |
| Level of Study: |
Graduate |
| Course Coordinator: |
|
| Offered Semester: |
Fall and Spring Semesters. |
Course Content
The course is divided into two parts: Hidden Marko models (HMMs) and methods for handling so called big data sets that fall under the heading of machine learning. First part covers model specification, parameter estimation (maximum likelihood using numerical maximization and the EM algorithm), model selection and checking, forecasting, local and global decoding (the Viterbi algorithm), state prediction. Several non-standard HMMs, HMM approximations to hidden semi-Markov models and to continuous state-space processes. Topics covered include supervised and unsupervised learning using various of methods of clustering, dimension reduction, tree-based and non-parametric regression.
