|
| Udgave: |
Forår 2013 NAT |
| Point: |
7,5 |
| Blokstruktur: |
4. blok |
| Skemagruppe: |
C |
| Fagområde: |
dat |
| Varighed: |
8 weeks
|
| Omfang: |
20 hours per week |
| Institutter: |
Department of Computer Science |
| Uddannelsesdel: |
Kandidat niveau |
| Kontaktpersoner: |
Kim Steenstrup Pedersen. Telefon 35 32 14 55
E-mail: kimstp@diku.dk |
| Andre undervisere: |
Christian Igel,DIKU
Thomas Hamelryck, Bioinformatics
|
| Skema- oplysninger: |
 Vis skema for kurset
Samlet oversigt over tid og sted for alle kurser inden for Lektionsplan for Det Naturvidenskabelige Fakultet Forår 2013 NAT |
| Undervisnings- form: |
Lectures and project supervision |
| Formål: |
The purpose of this course is to introduce the student to selected advanced topics in stochastic and deterministic modeling and analysis of real world sampled data. The techniques taught are widely applicable within computer science, bioinformatics, and E-Science. Application examples include general machine learning, image analysis, computer vision and biology. The course will be evaluated with a small written project within one of the advanced topics.
|
| Indhold: |
Foundation of stochastic modeling: The core part of the course covers a selection of topics such as:
• Graphical models
• Learning and inference
• Hidden Markov models (HMM)
• Markov models
• Markov random fields
A selection of image analysis topics will be covered such as:
• Kalman and particle filtering
• Visual tracking and Visual SLAM
• Motion analysis and optical flow
• Texture and image models with applications (texture synthesis and analysis, denoising, inpainting)
• Interest point detectors and image descriptors and applications
• Techniques for object recognition and image retrieval
• Shape modelling with applications
A selection of bioinformatics topics will be covered such as:
• Graphical models for structural bioinformatics
• HMMs for biological sequence analysis
• Stochastic context-free grammars for RNA structure
• Probabilistic models for expression analysis a
• Algorithms for large scale genomic mapping
|
| Målbeskrivelse: |
At course completion, the student should be able to:
1. Recognize and describe possible applications of selected stochastic and deterministic data models and analysis methods.
2. Explain, contrast and apply selected data representations.
3. Explain and contrast static and dynamic data models and their applications.
4. Apply static and dynamic data models within appropriate applications.
5. Implement selected methods and models. |
| Lærebøger: |
The course will be mainly based on selected research papers |
| Tilmelding: |
November 15 to December 1, 2012, via KUnet, www.kunet.dk |
| Faglige forudsætninger: |
Students must know the content of either the courses “Statistical Methods for Machine Learning”, “Machine Learning for Pattern Recognition” or similar.
Informally, the students are expected to have a mature and operational mathematical knowledge. Knowledge of linear algebra, geometry, basic mathematical analysis, and basic statistics is relevant.
We also expect that the student is able to program in a language suitable for scientific modelling. |
| Eksamensform: |
Continuous written assignments. Grade: 7-point scale. Internal grading. Submission in Absalon.
Re-exam: Resubmission of written assignments. Grade: 7-point scale. Internal grading.
|
| Eksamen: |
Løbende evaluering.
Reeksamen: Genaflevering af opgaver d. 21. august 2013. |
| Kursus hjemmeside: |
 |
| Undervisnings- sprog: |
Kun engelsk
|
| Sidst redigeret: |
31/10-2012 |