Unit details

STAT4063 Computationally Intensive Methods in Statistics

Credit

6 points

Offering
(see Timetable)

If this unit does not have an online alternative, then students who are presently unable to enter Western Australia and whose studies would be delayed by an inability to complete this unit, should contact the unit coordinator (details given on this page) to ascertain, on an individual case-by-case basis, if alternate arrangements can be made to support their study in this unit.

Availability	Location	Mode
Semester 2	UWA (Perth)	Face to face Predominantly face-to-face. On campus attendance required to complete this unit. May have accompanying resources online.

Details for undergraduate courses

• Level 4 elective
• Honours option in Mathematics and Statistics [Bachelor of Science (Honours)]

Content

The explosion in power of computers over recent times is changing the face of statistical science. Once upon a time, intuitively attractive statistical procedures had to be consigned to the waste bin if they led to unfathomable mathematical complications, or required masses of intricate calculations for their practical implementation. Now such difficulties can often be circumvented by using computer simulation and number crunching. This has led to the development and widespread use of many new statistical tools including the bootstrap, Markov Chain Monte Carlo methods and non-parametric kernel smoothing methods. Computer intensive statistical methods are not only in general use by statisticians, but are also applied by quantitative researchers in the life sciences, medicine and biological science, social sciences and business.

This unit gives a broad coverage of computer intensive methods with numerous applied examples, together with the underlying general concepts and basic theory. Particular emphasis is placed on the use of these methods in real statistical applications. Data sets are analysed using the statistical package R. Topics are selected from the following: simulation and Monte Carlo, bootstrap methods, Markov Chain Monte Carlo (MCMC) methods and Bayesian inference, non-parametric kernel smoothing methods, and statistical/machine learning.

Outcomes

Students are able to (1) appropriately apply computationally intensive statistical techniques in simulation studies and to real-world problems; (2) extend their knowledge of computational techniques in general, but statistical computing techniques in particular, and adapt known solutions to different situations; and (3) present results in a logical and coherent fashion and communicate effectively with others.

Assessment

Indicative assessments in this unit are as follows: (1) assignments and (2) a final examination. Further information is available in the unit outline.

Supplementary assessment is not available in this unit.

Unit Coordinator(s)

Dr Edward Cripps

Unit rules

Prerequisites:

STAT3062 Statistical Science or STAT2401 Analysis of Experiments

Advisable prior study:

STAT2402 Analysis of Observations (ID 389) or STAT2062 Fundamentals of Probability with Applications (ID 5019)

Contact hours

3 hours per week