|
Semester 1 2005:
Semester 2 2005:
 Postgraduate Opportunities 2005
|
Postgraduate Courses for 2005The courses listed on the left-hand panel are those we currently plan to run in 2005. Changing circumstances may, however, force some changes to these plans. Semester 1 2005STATS 708 Topics in Statistical Education
Statistics education is an exciting new field of research. This course covers a wide range of New Zealand and international research in statistics education at the school and tertiary level. The focus will be on how probability, statistical investigations, exploratory data analysis and statistical literacy can be effectively taught and learned. Issues involved in learning statistics will be examined through research literature discussions, practical applications for the classroom including computer activities, and lectures. Topics studied include: Probability - decision making under uncertainty, concepts of randomness, combinatoric reasoning; Statistical Reasoning - students' understanding of specific statistics content such as averages, distributions, inference; Statistical Investigations- critical teaching and learning barriers; Statistics and Technology- data analysis software, learning statistical concepts using technology; Interpreting Statistical Information- graphicacy theory, statistical literacy; Statistical Thinking- key ideas underlying data-based enquiry; Statistics in the Curriculum - current conceptual framework including assessment. By the end of STATS 708 all students should have aimed to produce a scholarly paper on one aspect of statistics education. This will involve a literature review on that aspect; the writing of the paper to a publishable standard; and a presentation of that paper to an audience. In addition, students should be able to state, and argue coherently for, their perspective on statistics education: how and why specific teaching and learning strategies should be used; the implications of research for teaching and learning; the relationship between statistical learning and assessment; and so forth. STATS 708 will involve lectures, practical tasks, and discussion. Each session will require pre-reading of papers and/ or a practical exercise. Each participant will comment on their reading and/ or practical exercise for the week. STATS 723 Stochastic Methods in Finance
STATS 723 is a course in mathematical finance. We will use the techniques of probability and stochastic processes to study derivative securities that are traded in financial markets. The course is suitable for students of statistics, finance, accounting, economics, or mathematics. Topics studied include: Probability and background - Random variables, distributions, equivalent measures. Essential notions - financial derivatives, arbitrage, replicating portfolios, risk-neutral pricing, complete markets. Martingales. Mathematical finance in discrete time -equivalent martingale measures, Fundamental Theorem of Asset Pricing, Cox-Ross-Rubinstein model. Option pricing, including path-dependent options. Continuous-time stochastic processes - Brownian motion, stochastic integral, Ito's lemma, stochastic differential equations. Mathematical finance in continuous time - Fundamental Theorem of Asset Pricing in continuous time. The Black-Scholes world. Hedge sensitivities ("Greeks"). Further topics - a selection from: volatility estimation, volatility smiles; risk management and hedging, risk measures (VaR and CVaR); multi-factor and other models, Monte Carlo simulation; incomplete markets. STATS 726 Time Series
STATS 726 provides a general introduction to the theory of time series and prediction including stationary processes, moving average and autoregressive (ARIMA) models, modelling and estimation in the time domain, seasonal models, forecasting, spectral analysis and bivariate processes. This foundation course is particularly suitable for students in economics and finance, and in the engineering and physical sciences. Specific topics covered include: the basic theory of stationary processes; spectral or Fourier models; AR, MA and ARMA models; linear filtering; time series inference; prediction and other topics such as the sampling of continuous time processes. STATS 747 Statistical Methods in Marketing
STATS 747 discusses statistical issues that arise and statistical techniques that are used in market research. Topics studied include: stochastic models of brand choice and usage, conjoint analysis, sampling, weighting, imputation and variance estimation, advertising media models and marketing response models, and applications of generalised linear models, data mining and multivariate analysis in market research. STATS 760 A Survey of Modern Applied Statistics
STATS 760 is an overview course aimed at giving students an introduction to, and some practical experience with, many of the most important forms of statistical analysis. The emphasis is on breadth rather than depth. In general, students will study a subset of the areas introduced here in depth in the rest of their postgraduate programme. Topics studied may include: generalised linear models; robust statistics; nonlinear models; random and fixed-effects models; modern regression; survival analysis; multivariate analysis; tree-based methods; time series; spatial statistics; classification. STATS 761 Mixed Models
STATS 761 is a useful course for those contemplating a career in Medical Statistics. This course will describe statistical techniques for analysis of data from medical studies, with an emphasis on mixed modelling. This is predominantly a practical rather than theoretical course. Students will be expected to program in SAS and to be able to interpret the resulting output. Examples of SAS code will be given in the lecture notes and explained. Coursework will involve using SAS to analyse real data sets. Topics studied include: Use of SAS for analysing medical data, with applications in epidemiology. Analysis of multicentre trials (random effects models), repeated measures data (covariance pattern and random coefficient models), and matched case-control studies. Use of generalized linear models in medical statistics. STATS 770 Introduction to Medical Statistics
STATS 770 provides a broad introduction to ideas of importance in medical statistics, such as measures of disease risk, types of disease, diagnosis, clinical and epidemiological studies, and reviews of common methodologies. Topics studied include: Topics are covered under three general headings: treatment, risk and diagnosis. Treatment will cover clinical trial methodology; study design; sample size; and analysis. Risk will cover statistical measures of risk in epidemiology and their estimation; data sources; analysis of mortality data; epidemiological study designs. Diagnosis will cover ideas about the nature of disease, statistical and probabilistic approaches to diagnosis; diagnostic test theory; screening for disease; logistic and discriminant classifying models; other methods of classification including tree models. STATS 775 Design of Ecological Experiments
Systems in nature have intrinsic spatial and temporal variability, which means that most studies of natural systems must involve complex experimental designs. STATS 775 covers the fundamental considerations in the design and analysis of experiments in ecological, biological and environmental scientific work. The emphasis is on linking the design with an appropriate statistical analysis. The course is particularly focused on hypothesis-testing methods and the interpretation of statistical results for complex experimental designs commonly used or encountered in these fields. This is a practical course, which emphasises statistics in the context of particular ecological and biological studies and examples. There are no explicit pre-requisites, but familiarity with the material in STATS 340 is an advantage. Topics studied include: The logical use of statistics and tests in ecological and environmental research; Replication and pseudo-replication for experiments in natural systems; Controls, sampling strategies and scales of observation in spatially and temporally variable systems; Factorial designs, nested hierarchies and mixed models; Fixed versus random factors in experiments and the consequences for analysis and interpretation; Calculating expected mean squares and constructing appropriate F-ratios for terms in complex designs in analysis of variance; How to interpret significant interactions and their scientific value; Asymmetric designs, contrasts and a posteriori tests; Experimental designs for detecting environmental impacts, including repeated measures, BACI and beyond BACI; Permutation tests, bootstrapping and Monte Carlo sampling methods for complex designs and misbehaving data. STATS 782 Statistical Computing
STATS 782 is an introduction to the computer as a tool for the professional statistician. The course will provide an introduction to: computers and their architecture; non-numerical computing including data manipulation; text processing and typesetting; and numerical computing using statistical packages and traditional computer languages. STATS 784 Statistical Data Mining
STATS 784 will discuss the nature of data mining and a selection of topics from: the classification problem, regression and decision trees, neural networks, fraud detection, data cleaning. Semester 2 2005STATS 724 Operations Research
STATS 724 is a course about stochastic modeling and optimization that follows on from STATS 320 and STATS 325. Students should have done either STATS 325 or STATS 720. Students who have done very well in STATS 320 may also be allowed into the course- please see Ilze Ziedins if you fall into this category. The course will cover a range of stochastic models, giving both theory and some discussion of applications. The emphasis will be on Markov models. Some queueing theory will be presented as part of the applied section. Some computing experience (a working knowledge of R or Splus or Matlab) will be useful. Topics covered may include: Markov processes in continuous time - definitions, Q-matrix, recurrence, stationary and limiting distributions, migration processes, reversibility. Queueing theory, including material on queueing networks. A key component of the course this year will be Markov decision processes and optimization. If time permits, the course may include some additional material on renewal theory and semi-Markov processes. STATS 730 Statistical Inference
STATS 730 gives you general-purpose skills to model real data, using likelihood-based statistical inference. It begins by looking at how likelihood is used by frequentist and Bayesian inference, and uses statistical brain-teasers to demonstrate the difference between these two paradigms. Focus then shifts to establishing the properties of maximum likelihood inference within the frequentist paradigm. Maximum likelihood is then applied in a wide variety of settings with examples in both R and SAS. (Students may chose either of these languages for their homework.) The course concludes by looking at extensions of maximum likelihood for models with nuisance parameters, including quasi-likelihood, conditional likelihood, and mixture models. STATS 730 provides the tools and skills used by many other graduate courses on offer in this department, and it gives exposure to statistical programming in both R and SAS. To understand the theory it is necessary that students are concurrently enrolled in (or have already completed) STATS 310. Students are also expected to have a good grasp of second-year calculus: multiple integrals, partial derivatives and Taylor series expansions will be used in this course. STATS 731 Bayesian Inference
STATS 731 is an introductory course in Bayesian inference starting from first principles with major emphasis on Bayesian methods in applied data analysis. The Bayesian approach is based on a different paradigm than the classical frequentist approach to statistical inference that is traditionally taught in undergraduate courses. Over the last decade, the Bayesian approach has revolutionised many areas of applied statistics such as biometrics, econometrics, market research, statistical ecology and physics. Although the Bayesian approach dates back to the 18th century, its rise and enormous popularity today is due to the advances made in Bayesian computation through computer-intensive simulation methods. Knowledge of Bayesian procedures and software packages will become indispensable for any career in Statistics. The students will be introduced to the software packages FirstBayes and BUGS. Topics covered include: the Bayesian approach, conjugate distributions, using FirstBayes, prior distributions, simulation methods, Markov chain Monte Carlo methods, using the BUGS software, and applications to data analysis. STATS 740 Sample Surveys
STATS 740 is a course in the design, management and analysis of sample surveys, with particular emphasis on surveys in the health sector. Topics studied include: Revision of statistical aspects of sampling. Preparing surveys. Research entry: problem selection, sponsorship and collaboration. Research design: methodology and data collection; Issues of sample design and sample selection. Conducting surveys: Questionnaires and questions; Non-sampling issues; Project management; Maintaining data quality. Concluding surveys: Analysis; Dissemination. Other types of Surveys. BIOSCI 744 Advanced Bioinformatics 2
[This course was taught in 2003 as STATS 771 Topic in Biostatistics A (Bioinformatics)] This course gives an introduction to structural, functional and comparative genomic analysis, and the analysis of microarray and gene expression data. It will also survey recent developments in bioinformatics. Microarray technology allows the simultaneous measurement of gene expression (the level of activation of a gene) for thousands of genes in a single experiment. This powerful technology has become incredibly popular in genetics research, as it is able to provide researchers with huge amounts of information about genetic activity. The amount of data that is generated, however, poses problems in terms of analysis (data sets can easily have tens, or even hundreds of thousands of data points), making statistical methodology useful in processing such large amounts of data. Statistical gene-mapping involves detection of associations between traits (such as diseases) and genetic markers for the purpose of determining which genes are involved in influencing trait values. Improving technology has greatly increased the availability of genetic markers and decreased the costs of genotyping. The resulting data explosion presents challenges which include how to account for correlation between markers, and multiple testing issues. This course aims to teach students how to design and analyze gene expression and gene-mapping experiments, skills which are in high demand by biotechnology, bioinformatics and pharmaceutical companies throughout the world. The skills learned in this course will also be applicable to other problems involving large data sets, such as data-mining and proteomics. Topics covered include: basic biology as it relates to microarray analysis and gene-mapping, linkage-disequilibrium gene-mapping, microarray technology and methodology, multiple comparisons procedures, clustering, and non-parametric techniques. This course is intended for fourth year statistics students and students pursuing a B.Sc.(Hons) degree in Bioinformatics, but will also be accessible to biology students with some knowledge of statistics. STATS 764 Analysis of Failure Time Data
STATS 764 will cover topics in the theory and analysis of time to event data, otherwise known as survival data, or failure time data. Here the variable of primary interest is the time elapsed until some event occurs (e.g. death of a person, or failure of a component). Such data are often subject to censoring and truncation and thus require specialised methods of analysis. Two of the most important areas of application are the health sciences and industrial testing (where the term reliability is often used). Topics studied include: Censoring and truncation; commonly used distributions; nonparametric estimation (including Kaplan-Meier estimators); exploring censored data; parametric regression models; semiparametric regression models emphasizing the proportional hazards model; time-dependent covariates; rank tests; and multivariate survival times. STATS 771 Topics in Biostatistics 1
In 2006, this number will be used to enrol 2nd-year Masters students doing the Bioinformatics course (see BIOSCI 744) for the MSc in Medical Statistics. All others will be enrolled in BIOSCI 744. STATS 773 Design and Analysis of Clinical Trials
STATS 773 is directed at students interested in medical statistics and covers the design and analysis of clinical trials. The emphasis will be on practice rather than theory, and the course will cover some of the management issues arising in clinical trials such as ensuring data integrity and confidentiality as well as the statistical issues. An extensive case study will be discussed. STATS 780 Statistical Consulting
This course is designed to give students training in statistical consulting, i.e., applying statistical methods to practical research problems in other disciplines. The first half of STATS 780 mostly consists of lectures and assignments. The second half of the course focuses on a group project; together with a member of staff of the Statistics Department, small groups of students will discuss with researchers and other clients the design and analysis of quantitative investigations. The course is not designed to teach students specific statistical techniques, but focuses instead on the other skills required for statistical consulting. Students will be assessed on their ability to present both written and verbal reports. This will include a seminar based on the group project to be delivered to the client and members of the Statistics Department. It is assumed that students will have passed STATS 330 and STATS 340. There is no final exam. STATS 782 Statistical Computing
STATS 782 is an introduction to the computer as a tool for the professional statistician. The course will provide an introduction to: computers and their architecture; non-numerical computing including data manipulation; text processing and typesetting; and numerical computing using statistical packages and traditional computer languages. STATS 787 Statistical Graphics (Topics in Statistical Computing)
STATS 787 will cover some important topics in statistical graphics, including: the production of dynamic and interactive graphics for data exploration; the production of high-quality static graphics for presentation; human perception and its impact on the design of graphics; the selection of appropriate graphics formats. The principal software tool for this course will be R. |
