Simulation Methods in Statistics

6 EC

Semester 1, period 1, 2

5374SIMS6Y

Owner Master Mathematics
Coordinator dr. A.J. van Es
Part of Master Mathematics, year 1Master Stochastics and Financial Mathematics, year 1

Course manual 2017/2018

Course content

Nowadays simulation methods based on random number generation on powerful computers play an important role in statistics. We highlight two methodologies, the bootstrap and Monte Carlo Markov chain simulation. The bootstrap method has been introduced in 1977 by Bradley Efron. This is a useful, generally applicable, but computationally intensive, method to construct, for instance, confidence intervals. The basic idea of the method is resampling from the original data. The naive bootstrap, paramatric bootstrap and smooth bootstrap shall be discussed.

By running a computer simulated Markov chain for a suitably long time we can generate observations from a distribution close to the stationary distribution of the Markov chain. By choosing suitable transition probabilities practically any distribution can be simulated in this way. We will discuss the Gibbs and Metropolis Hastings algorithms, the basic algorithms for this kind of simulation, as well some of their refinements, bearing in mind the relevance for statistics.

Study materials

Syllabus

  • Lecture notes: The Bootstrap, Bert van Es and Hein Putter

  • Monte Carlo Markov Chain Simulation, Bert van Es

Objectives

The student should

  • be able to explain the bootstrap method;
  • be able to apply the bootstrap method in simple theoretical exercises
  • be able to explain and apply several standard computer simulation methods like inversion and acceptance/rejection sampling;
  • be able to explain the MCMC method, in general and in certain applications;
  • be able to apply the MCMC method on the computer in practical situations.

Teaching methods

  • Lecture
  • Exercises

The theory is presented during the lectures. The exercises serve as a way to get a grip on the theory.

Learning activities

Activity

Number of hours

Lectures

28

Self study

28

Exercises

42

 

Attendance

The programme does not have requirements concerning attendance (OER-B).

Assessment

Item and weight Details

Final grade

25%

Exercises

75%

Final oral exam

The course consists of two parts. The first part, The Bootstrap, has most of the exercises. The second part, MCMC, has less exercises.

The oral exam concerns the second part, MCMC, only.

Fraud and plagiarism

The 'Regulations governing fraud and plagiarism for UvA students' applies to this course. This will be monitored carefully. Upon suspicion of fraud or plagiarism the Examinations Board of the programme will be informed. For the 'Regulations governing fraud and plagiarism for UvA students' see: www.student.uva.nl

Course structure

Weeknummer Onderwerpen Studiestof
1  Introduction. Empirical distribution function.  Bootstrap: Chapters 1 and 2
2

 Different bootstrap methods. Bias and variance bootstrap estimators.

Bootstrapping by computer simulation.

 Bootstrap: Chapters 3 and 4 (skip Section 4.2)
3  The Jackknife method.  Bootstrap; Chapter 5
4  Proof that the bootstrap works for the sample mean.  Bootstrap: Skip Chapter 6, Do Chapter 7
5    Free
6  Example where the bootstrap fails and the m out on n bootstrap as a remedy.  Bootstrap: Chapter 8
7  Accuracy of the bootstrap. Studentized bootstrap, Kernel smoothing.  Bootstrap: Chapter 9 and Appendix C
8    Free
9 The smooth bootstrap: quantile  Bootstrap: Chapter 10
10 Introduction and standard simulation methods  MCMC: Chapters 1, 2
11 Brief review od standard Markov Chains, Chapter 3, and MCM chains, Chapter 4. Gibbs and Metropolis Hastings samplers.  MCMC: skip Chapter 3. Do Chapter 4
12 Chapter 5. One dimensional random number generation by Metropolis Hastings chains. Independent and random walk MH.  MCMC: Chapter 5
13 Chapter 6. MCMC for a Bayesian analysis of a contingency table.  MCMC: Chapter 6
14 Chapter 7. An example in Finance. Estimation of the parameters of a geometric Brownian motion from asset and option prices.  MCMC: Chapter 7
15    
16    

 

Timetable

The schedule for this course is published on DataNose.

Additional information

Recommended prior knowledge: Measure Theoretic Probability, Basic statistics.

Contact information

Coordinator

  • dr. A.J. van Es