 |
Discrete Stochastic Modelling covers approximately half the material contained in the Institute and Faculty of Actuaries' syllabus for subject 103 (Stochastic Modelling). The remaining half is covered by Continuous Stochastic Modelling, which will be taught in Term 1 next year.
Although there will be a slant in favour of actuarial models, not all models will be presented in direct connection with actuarial applications.
In weeks 7, 9 and 11 there will be lab sessions. Two are associated with the Simulation element of the course, the third with Time Series. We shall be using Excel to perform simulations and Minitab for the Time Series analysis.
I anticipate producing two coursework sheets, in weeks 5 and 8. These will be produced automatically by the coursework generator. The hand-in date will be clearly marked in each case.
A penalty of 2% of the maximum mark per day (including weekends) will be applied to late coursework. Coursework handed in after the solutions are published will receive no credit.
There will be some element of personalisation about all coursework sheets.
From time to time there will be a question set during a lecture for you to solve on the spot, with 2 or 3 coursework marks available for a correct answer. A maximum of 20% of all coursework marks will be available for the cumulative score on all "two-pointers".
A two-hour examination will take place in May. As in previous years, there will be five questions, of which you will have to answer three.
The transparencies used in lectures will be published on the course web page, http://www.staff.city.ac.uk/r.j.gerrard/courses/2dsm/.
Other materials will also be indexed from that page, when they become available. You are advised to make use of the Bookmark facility in Netscape Communicator (called Favorites in Internet Explorer) so that you can find the page again next time.
The Institute and Faculty of Actuaries' Core Reading for Subject 103 comes closest to covering the material included in this course. See unit 1, part of unit 2, unit 3, unit 5 and unit 7.
Most books in this area cover either Markov Chains or Time Series but not both. A recent book on Markov chains is by Norris (Markov Chains, CUP, paperback 1998 under £20 from Amazon), or try Grimmett and Stirzaker (Probability and Random Processes, OUP, 3rd edition paperback 2001 approx £30 from Amazon) or Ross (Introduction to Probability Models, Academic Press, 8th edition paperback 2003 approx £35). For Time Series I like Chatfield (The Analysis of Time Series, CRC, 6th edition paperback 2003) but Diggle (Time Series: A Biostatistical Introduction, 2000: hard to find) is particularly good for the practical aspects.
The Library has a number of stochastic processes texts - some old, some new. If you find a point unclear, you are welcome to try one.
The Web has a small number of relevant courses online. The course web page will eventually include a list of links to the various sites, with specific references to individual pages where they offer something unusual such as an interactive facility. (You may need to have Java enabled in your browser to take advantage of them.)
From 2005 Subject CT4 will replace Subject 103. In order to qualify for an exemption from subject CT4 you will need to have passed both 103 and 104.
Introduction.
Be warned that the various sections are not of equal length or of equal difficulty.
|