Important Dates and Deadlines
Project deadline will be 3 days before the exam date.
There will be two exam dates one at the beginning of June and the other on the last day of the exam session.
Lectures
- Introductory Seminar (10 February, Sven Laur,pdf)
- Graph construction and visualisation (17 February, Balaji Rajashekar, Chapter 3, Attach:seminar-2 Δ)
- Theoretical overview
- Case study for network construction (Internet2 data or co-expression data)
- Case study for network visualisation with Cytoscape
- Local and global characteristics (3 March and 10 March, Chapter 4, Markko Merzin and Svitlana Vakulenko, pdf1 pdf2, code)
- Theoretical overview of local characteristics and outlier nodes
- Case study of local characteristics: histograms, correlations, changes in time, interpretation of results.
- Theoretical overview of global statistics: centrality, connectivity, local density
- Experimental study of global characteristics: How local changes (deletions and insertions) alter global statistics
- Theoretical overview of partitioning and clustering algorithms
- Case study of partitioning algorithms: use visualisation and descriptive statistics to show the differences between various partitioning algorithms
- Sampling and Estimation in Network Graphs (17 March, Chapter 5, Anastassia Semjonova and Roman Tekhov, pdf)
- Theoretical overview: sampling strategies, bias and specific estimators
- Case study of sampling: how well sampling strategy works with restrictions, e.g. Facebook sampling and network crawling
- Classical Probabilistic Models for graphs (24 March, 153-168, Abel Armas pdf)
- Theoretical overview of Erdős–Rényi model for random graphs
- Case study of Erdős–Rényi model: use descriptive analysis to study its properties and compare it with real world graphs.
- Case study of Erdős–Rényi model: Visualise empiric phase transitions for the graph properties.
- Additional materials:
- Small-World Models and Network Growth Models (31 March, 169-180, Anastassia Semjonova and Roman Tekhov, pdf)
- Theoretical overview
- Case study of small-world and network growth models: how these models differ from Erdős–Rényi and real world graphs
- Additional materials:
- Exponential Graph Models (7 April, 180-194, Raivo Kolde ja Karl Potisepp, pdf R-code)
- Theoretical overview
- Case study of exponential graph models: descriptive analysis and fitting the model to real world network
- Additional materials:
- http://www.sna.unimelb.edu.au/publications/ERGM11.1.pdf
- ergm: A Package to Fit, Simulate and Diagnose Exponential-Family Models for Networks
- Specification of Exponential-Family Random Graph Models: Terms and Computational Aspects
- statnet: Software Tools for the Representation, Visualization, Analysis and Simulation of Network Data
- Basic Methods for Link Prediction (14 April, 197-207, Riivo Kikas and Oskar Gross, pdf)
- Theoretical overview of logistic regression and model validation methods
- Case study: building and evaluating logistic regression predictors on real data
- Use standard machine learning and statistic textbooks as a supplementary material if needed
- SVM-s and Kernel methods for Graphs (28 April, 257-271, Aleksandr Tkatšenko, pdf)
- Theoretical overview
- A case study on real data
- Use standard machine learning textbooks such as Kernel Methods as a supplementary material if needed
- Inference and Prediction with Markov Random Fields (5 April, 245-257, Raivo Kolde ja Karl Potisepp, pdf)
- Theoretical overview
- A case study on real data
- Modeling Dynamic Processes in Graphs (12 May, 271-281, Riivo Kikas and Oskar Gross, pdf)
- Theoretical overview
- A case study on real network data
- Tree reconstruction algorithms (19 May, 223-241, Lauri Eskor and Markus Läll, pdf )
- Theoretical overview
- A case study on real network data
- Gravity Models (26 May, 285-297, Lauri Eskor and Markus Läll, pdf)
- Theoretical overview
- A case study on real network data
Other Topics Not Discussed in the Seminar
- Traffic Estimations in Networks (19 May, 297-316, ??)
- Theoretical overview
- A case study on real network data
- Estimations of Network Flow Costs (19 May, 317-328, ??)
- Theoretical overview
- A case study on real network data
- Correlation Networks Revisited (???, 207-223, ??)
- Theoretical overview of Gaussian Graphical models
- A case study on real biological co-expression data
- This is a really complex topic about graphical models that deserves a separate course