MATS255 Reading Assignment
This is a reading exercise where a group of 3 students (recommended) or 2 students (also possible)
reads and presents to others a selected research article related to the course theme. Based on the article,
the group will give a 15–20 min presentation, where the following four key points will be discussed:
- Objective (what is the goal of the research).
- Motivation (why is the research worth carrying out).
- Results (what are the main results obtained).
- Evaluation (what is your own estimate of the usefulness of the results).
Reading seminar
The presentations will take place in a reading seminar on Mon 12 Dec at 14:15–18:00 in room MaD302.
Presentations
- Pekka Aalto & Nanni Koski:
Chip-firing games on graphs
A. Björner, L. Lovasz, P. W. Shor
European Journal of Combinatorics 12(4):283–291, 1991
- Sergey Chernov, Fedor Chernogorov, Evgeni Kuzmichev:
Simulated annealing
Olle Häggström
Chapter 13 in Finite Markov Chains and Algorithmic Applications. Cambridge University Press 2002
- Jouni Helske & Tuomas Kukko:
Modeling curling as a Markov process
Kent J. Kostuk, Keith A. Willoughby, Anton B. H. Saedt
European Journal of Operations Research 133:557–565, 2001
- Juho Hynynen & Aleksi Koskikallio:
Tilastollisia menetelmiä dynaamisten potilaspopulaatioiden mallintamiseen;
Tapahtumahistoria-analyysia hoitoilmoitusrekisterin skitsofreenikoille
Reijo Sund
Aiheita-monistesarja 26, Stakes, 2000
- Ari Kaihlanen & Santtu Salmi:
Lévy process simulation by stochastic step functions
Torquil Macdonald Sørensen, Fred Espen Benth
arXiv:1110.2367
- Benedict Kalus, Tuomo Sipola & Mari Supperi:
A Markov model for HIV disease progression including the effect of HIV diagnosis and treatment: Application to AIDS prediction in England and Wales
Odd O. Aalen, Vernon T. Farewell, Daniela De Angelis, Nicholas E. Day, O. Nöel Gill
Statistics in Medicine 16:2191–2210, 1997
- Ilari Korhonen:
Password strength : An empirical analysis
Matteo Dell'Amico, Pietro Michiardi, Yves Roudier
Infocom'10, 29th Conference on Computer Communications, 2010
- Mikko Kuronen & Lasse Liukkonen:
On the stochastic matrices associated with certain queuing processes
F. G. Foster
The Annals of Mathematical Statistics 24(3):355–360, 1953
- Jia Liu & Anna-Kaisa Ylitalo:
Propp–Wilson algorithm
Olle Häggström
Chapter 10 in Finite Markov Chains and Algorithmic Applications. Cambridge University Press 2002
- Eero Neijonen & Juho Nuutinen:
A Markov chain Fibonacci model
Myron Hlynka, Tolulope Sajobi
Missouri Journal of Mathematical Sciences 20(3):186–199, 2008
Guidelines
- The presentation is graded at the scale 0–3. Each member in a group receives the same grade.
- The main criterion for evaluating the presentation is how well the audience can get an overview of the aforementioned four key points of the article.
- 20 min is a very short time to present an article. In such short time there is usually no hope of transmitting all wisdom contained in a research article to an audience who are
not specialists in the field. Don't try to do this. Try to focus on the essential points of the article, or select to treat only a part of a longer paper.
- The style of presenting is free: you can use the blackboard, the beamer, or anything you find suitable.
- In a short presentation there is no need (actually, no use) to go through any detailed proofs or technical details.
- You do need to discuss all the aforementioned four key points.
- Don't choose too difficult article for reading, it will be too hard to present in a short time to the others.
If you cannot participate in the reading seminar on Mon 12 Dec, you may replace the presentation by writing an essay of 5–8 pages where you discuss the
same points as you would in the presentation. The essay will be graded at the scale 0–3, and the main criterion for evaluating the presentation is how well a randomly
selected fellow student from the course would be able get an overview of the four key points of your chosen article. The essay should be written alone.
Choosing an article
You have full freedom in proposing a research article, but your proposal must be confirmed by the teacher before the seminar.
When you have found a suitable article, send email to the teacher and ask it to be confirmed. Some guidelines:
- The article must have something to do with Markov processes.
- You can also choose to present a selected section of a longer research article
- Instead of a journal article, you can also present a book chapter.
- The article may be very theoretical or very applied. Applied articles are usually easier to present in a short time.
Some suggested topics:
- A Markov model related to wireless data networks, signal processing, or such
- Markov mixing rates in card shuffling
- A random walk model on a (random or nonrandom) graph
- A Markov model related to population biology or epidemiology
- A Markov model related to economics, finance, or insurance
- A Markov model related to game theory
- Birth–death processes
- Quasistationary birth–death processes
- Markov chains and harmonic analysis
- Markov chain ergodic theorem
- Approximate counting (e.g. Chapter 9 in Häggström's book)
- Coupling from the past (e.g. Chapter 10 in Häggström's book)
- Simulated annealing (e.g. Chapter 13 in Häggström's book)
- Convergence of infinite-state Markov chains using Lyapunov functions
Possible papers:
Sources for research articles:
2011-12-10 Lasse Leskelä