Past Student Probability and Statistics Seminars
2008/2009 Academic year2007/2008 Academic year
2006/2007 Academic year
2005/2006 Academic year
2008/2009 Academic year
-
8 OctoberNo seminar this week due to MAGIC lauch event
200815 OctoberError-in-Variable Problem
2008
A.Althubaiti (The University of Manchester)
4.00pm - Frank Adams Room 2, Alan Turing BuildingAbstractNone
22 OctoberA Brief Introduction to Kriging Methodology
2008
S.Das (The University of Manchester)
4.00pm - Frank Adams Room 2, Alan Turing BuildingAbstractNone
29 OctoberThe Truth About the Real World - Job Interviews and Assessment Centres
2008
E.Jones (The University of Manchester)
4.00pm - Frank Adams Room 2, Alan Turing BuildingAbstractNone
5 NovemberNo seminar this week - Reading week
200812 NovemberVariable Selection using h-likelihood
2008
C.Charalambous (The University of Manchester)
4.00pm - Frank Adams Room 2, Alan Turing BuildingAbstractNone
19 NovemberModelling Externalities: Environmental Economics
2008
J.Sexton (The University of Manchester)
4.00pm - Frank Adams Room 2, Alan Turing BuildingAbstractNone
26 NovemberDead Rats - An Exploration of the Financial Crisis
2008
J.Du Toit (The University of Manchester)
4.00pm - Frank Adams Room 2, Alan Turing BuildingAbstractThe rate race is over - the rats are dead! Long live the rats! Everyone is aware that in the last few months the world has experienced a rather dramatic financial crisis. This talk will explore some of the causes of that crisis. I am no economist and do not claim to understand the whole situation, however I will present my own view of the main causes of the mess and try to show how the pieces fit together. Along the way we will meet blue collar workers, loan pimps, reckless presidents, fat bankers and derivative securities. WARNING: This talk is rated Certificate 18 as it contains scenes of strong language and prejudice against greedy capitalist bankers. Viewer discretion is advised.
3 DecemberNo seminar this week
200810 DecemberPrincipal components analysis and functional data.
2008
T.Sheppard (The University of Manchester)
4.00pm - Frank Adams Room 2, Alan Turing BuildingAbstractNone
17 DecemberMean, Quantile and Variance Regression.
2008
C.Kou (The University of Manchester)
4.00pm - Frank Adams Room 2, Alan Turing BuildingAbstractNone
2nd Semester 2008/2009
-
4 FebruaryNo seminar this week.
2009 -
11 FebruaryNon-parametric Regression.
2009
A.Vrahimis (The University of Manchester)
4.00pm - Frank Adams Room 2, Alan Turing BuildingAbstractNone
18 FebruarySpectral analysis and its applications.
2009
Jie Cheng (The University of Manchester)
4.00pm - Frank Adams Room 2, Alan Turing BuildingAbstractNone
25th FebruaryNo seminar this week.
20094 MarchNo seminar this week.
200911 MarchTBA
2008
Y.Chen (The University of Manchester)
4.00pm - Frank Adams Room 2, Alan Turing BuildingAbstractNone
18 MarchVariable Selection methods in Longitudinal Data Analysis.
2008
H.Chao (The University of Manchester)
4.00pm - Frank Adams Room 2, Alan Turing BuildingAbstractNone
25 MarchIntroduction to Missing Data.
2008
D.Li (The University of Manchester)
4.00pm - Frank Adams Room 2, Alan Turing BuildingAbstractNone
22 AprilTBA
2008
X.Gao (The University of Manchester)
4.00pm - Frank Adams Room 2, Alan Turing BuildingAbstractNone
29 AprilTBA
2008
J.Du Toit (The University of Manchester)
4.00pm - Frank Adams Room 2, Alan Turing BuildingAbstractNone
6 MayTBA
2008
T.Xu (The University of Manchester)
4.00pm - Frank Adams Room 2, Alan Turing BuildingAbstractNone
13 MayTBA
2008
K.Liu (The University of Manchester)
4.00pm - Frank Adams Room 2, Alan Turing BuildingBack to Probability and Statistics Research SeminarsAbstractNone
2007/2008 Academic year
-
17 OctThe Essence of Kernel Density Estimation
2007
A. Vrahimis
4.00pm - Alan Turing - Frank Adams Room 2 -
24 OctSemiparametric Time-Varying Coefficient Mean-Covariance Modelling
2007
C. Kou
4.00pm - Alan Turing - G.207 -
31 OctSimple Random Walks: The Ballot problem & Last Visits
2007
E. Jones
4.00pm - Alan Turing - Frank Adams Room 2 -
7 NovMaximum Entropy in Time Series
2007
B. Iqelan
4.00pm - Alan Turing - Frank Adams Room 2 -
14 NovExtended GEEs with Random Effects in Longitudinal Studies
2007
C. Huang
4.00pm - Alan Turing - Frank Adams Room 2 -
21 NovRandom Effects Covariance Modelling in Generalised Mixed Models
2007
E. Tan
4.00pm - Alan Turing - Frank Adams Room 2 -
28 NovFrom Condoms to Kolmogorov's Equations: Free Boundary Problems in Optimal Stopping
2007
J. Du Toit
4.00pm - Alan Turing - Frank Adams Room 2 -
5 DecIsotonic Regression and Longitudinal Data
2007
D. Li
4.00pm - Alan Turing - Frank Adams Room 2 -
12 DecOptimal Solution to Vehicle Routing Problem based on Tolerance
2007
S. Almoustafa
4.00pm - Alan Turing - Frank Adams Room 2 -
30 JanAn Introduction to Functional Data Analysis
2008
T. Sheppard
4.00pm - Alan Turing - Frank Adams Room 2 -
13 FebBorel-Cantelli Lemma
2008
Can a monkey be a famous author?
M. Savov
4.00pm - Alan Turing - G.207 -
5 MarSome MCMC concepts & their application to Integer-valued Time Series
2008
V. Enciso-Mora
4.00pm - Alan Turing - Frank Adams Room 2 -
12 MarDetection of Periodic Autocorrelations for Time Series data
2008
X. Gao
4.00pm - Alan Turing - Frank Adams Room 2 -
9 AprilProbability in option pricing & the British option
2008
F. Samee
4.00pm - Alan Turing - Frank Adams Room 2 -
16 AprilAn Introduction to Large Deviation Theory
2008
T. Xu
4.00pm - Alan Turing - Frank Adams Room 2 -
23 AprilSalamander mating data analysis
2008
Y. Chen
4.00pm - Alan Turing - Frank Adams Room 2 -
30 AprilAn introduction to the EM algorithm and its applications
2008
J. Cheng
4.00pm - Alan Turing - Frank Adams Room 2 -
7 MayLevy process and its application in finance
2008
M. Liu
4.00pm - Alan Turing - Frank Adams Room 2
Abstract
AbstractNone
AbstractNone
AbstractNone
AbstractNone
AbstractNone
AbstractThis week's talk will explore one of the most important results in probability theory - Kolmogorov's equations. These equations are particularly useful in solving applied problems and we will see how they are used in Optimal Stopping. These equations also have a deeper connection with stochastic calculus, which well consider if there is time.
We will also discuss condoms.
The talk should be accessible to anyone with some intuitive notion of what random variables and stochastic processes are, and what the distribution of a random variable is (roughly).
Knowledge of condoms is not essential.
AbstractNone
AbstractNone
Abstract
AbstractNone
AbstractNone
AbstractNone
AbstractNone
AbstractAbstractNone
AbstractNone
AbstractNone
2006/2007 Academic year
-
18 OctoberMultivariate Regression with Correlated Errors - An Analysis of Monthly Antartic Temperatures
2006
X. Gao
4.00pm - Ferranti C.18 -
8 NovemberModelling the Random Effects Covariance Matrix for Generalized Linear Mixed Models using the GEM Algorithm
2006
E. Tan
4.00pm - Ferranti C.18 -
15 NovemberRandomized QMC Approximation for Estimation in Generalized Linear Mixed Models
2006
E. Al-Eid
4.00pm - Ferranti C.18 -
6 DecemberBayesian Analysis for Joint Modelling of Longitudinal and Survival Data
2006
Y. Bao
4.00pm - Ferranti C.18 -
14 FebruaryUsing LMS methods for Reference Growth Charts
2007
T. Kecovevic
4.00pm - Ferranti C.18 -
7 MarchSmooth Random Effects Distribution in a Linear Mixed Model using QMC
2007
Y. Chen
4.00pm - Ferranti C.18 -
21 MarchAn Automatic Bandwidth Selection Method of the Bivariate Additive Model in the Presence of Interaction
2007
Y. Wang
4.00pm - Ferranti C.18 -
25 AprilAn Epidemic Model for Discrete Data in Dynamic Population
2007
Y. Wei
4.00pm - Ferranti C.18
AbstractNone
AbstractNone
AbstractNone
AbstractNone
AbstractNone
AbstractNone
AbstractNone
AbstractWe propose a stochastic epidemic model for discrete epidemic data, by allowing births and deaths in the population. The unknown parameters of interest are modelled in terms of the individuals' covariates. We consider to model birth and infection probabilities related to both of individuals' covariates and individuals' spatial location, whilst the removal and death are only individual dependent. We look at the problem when individual are captured at discrete time point, some of them are captured at every time point but some of them are not, which are called missing data. The estimation of the parameter of interest is achieved by using permutation for the missing data and MCMC techniques to explore the posterior distribution of parameters of interest. We also consider the model selection, and looking at the posterior model probabilities, which are obtained by using reversible jump Markov Chain Monte Carlo RJMCMC (Peter Green,1995). The method is applied to simulated data.
2005/2006 Academic year
-
24 MarStatistical analysis of a discretely observed dynamic epidemic model
2006
Y. Wei
10.00am - Ferranti C.18 -
28 AprilEfficient use of LaTeX and R in research
2006
G. Boshnakov
10.00am - Ferranti C.18 -
12 MayGeneralisations of Principal Components to Several Groups
2006
T. Sheppard
10.00am - Ferranti C.18 -
26 MayHistory and Background of Time Series
2006
J. Cheng
10.00am - Ferranti C.18
AbstractAvoiding the complication caused by dynamic populations, classical inference for infectious diseases assumes that the population is closed. Therefore it is assumed that the population of interest remains constant during the observational period. This is clearly unrealistic. We propose a stochastic epidemic model for discrete epidemic data which includes births and deaths into the population. Since a high proportion of the data concerning the epidemic is usually mixing there are problems with using classical statistical techniques. Instead, we treat the missing data as unknown parameters and MCMC is employed to estimate the parameters of interests. The model is applied to a Cowpox dataset.
AbstractNone
AbstractNone
AbstractNone
Further information
For further information contact
