STATS 760 - PowerPoint PPT Presentation

1 / 21
About This Presentation
Title:

STATS 760

Description:

... photocopier either in the City stats office (see Sharon Walker) or at ... The City Stats office has copies of Venables and Ripley available for photocopying ... – PowerPoint PPT presentation

Number of Views:32
Avg rating:3.0/5.0
Slides: 22
Provided by: Mat57
Category:
Tags: stats | stats

less

Transcript and Presenter's Notes

Title: STATS 760


1
STATS 760
  • Lecture 1

2
My contact details
  • Office (s)
  • Rm 228 City
  • 721332 Tamaki
  • Phone
  • Extn 88751 (City)
  • Extn 88749 (Tamaki)
  • Email lee_at_stat.auckland.ac.nz

3
Course Aims
  • To give you a survey of applied statistical
    techniques
  • To develop your skills in researching new
    techniques
  • To give you some practice in speaking and writing
    about statistics

4
Course Plan
  • There will be very few lectures!!!
  • Based around weekly one-on-one meetings
  • Each of you will independently study 5 topics,
    with help from me
  • In pairs, you will write a short paper on one of
    the topics and give a verbal presentation on the
    topic

5
Topics (first 5)
  • Linear Models
  • Glms/gams
  • Modern regression
  • Non-linear regression
  • Mixed models

6
Topics (Next 5)
  • Multivariate Analysis
  • Survival Analysis
  • Time Series Analysis
  • Spatial Statistics
  • Graphics/plotting

7
Journals and meetings
  • You will keep a journal in which you describe the
    things you have learnt, and give sample analyses.
  • At our weekly meeting, we will discuss what you
    have done, and discuss what is to be done next.
  • The final meeting will be an oral exam when I
    will ask you some questions to see what you have
    absorbed over the whole of your reading.

8
Assessment
9
How to approach a topic
  • For each topic, include in your journal
  • An appreciation of why the technique is important
  • The kinds of applied problems that the technique
    can solve
  • The types of data that call for the technique
  • Details of the models being fitted
  • The software (mainly R but sometimes SAS) that
    you need to implement the technique
  • How to interpret the output from the computer
    runs
  • Any diagnostic techniques that are important
  • How the technique relates to others (e.g. how to
    glms relate to linear models, GAMS to glms)

10
Resources
  • Use the following resources to research your
    topics
  • Modern Applied Statistics with S-Plus (MASS)
  • Other books (see reading list)
  • Encyclopedia of Statistics
  • Computer manuals
  • Google

11
Web pages
  • Course information
  • Notice board - check frequently!
  • Reading List
  • Assigned topics
  • R resources
  • Suggestions on writing papers and giving
    presentations

12
(No Transcript)
13
(No Transcript)
14
Lectures
  • Next week (Tuesday) an introductory lecture on
    linear models, and generalized linear models/
    generalized additive models
  • Other lectures as required, to be arranged if
    necessary I suggest a class meeting in about 3
    weeks to review procedures, make changes if
    necessary

15
Meetings(in my Tamaki office)
  • Tues 930
  • Tues 1000
  • Tues 1030
  • Tues 11
  • Tues 1130
  • Tues 400
  • Thurs 930
  • Thurs 1000
  • Thurs 1030
  • Thurs 1100
  • Thurs 1130
  • Thurs 200
  • Thurs 230
  • Thurs 300
  • Thurs 330
  • Thurs 400

Other meetings as required
16
Presentations and Papers
  • Choose a paper partner
  • Choose a topic
  • Email me by this Friday who your partner is and
    what topic you want to do
  • (Linear models is not an option)

17
Getting started Linear Models
  • Suggested topics (Linear models)
  • Review of 330
  • The Linear model
  • Model matrices
  • How R builds up the linear model from the model
    formula
  • More on prediction

18
Linear Models (ii)
  • Resources
  • Venables and Ripley Ch 6
  • STATS 330 lecture notes
  • (www.stat.auckland.ac.nz/lee/330
  • Reading list (000s of books on regression,
    linear models, see web page for recommended
    tiltles)

19
Getting started GLMs
  • Suggested topics
  • Exponential families
  • Maximum likelihood
  • Newton-Raphson algorithm
  • Iteratively reweighted least squares
  • Special cases (logistic/poisson/normal regression)

20
Graphics (ii)
  • Resources
  • Venables and Ripley Ch 4
  • STATS 330 lecture notes(trellis graphs)
  • (www.stat.auckland.ac.nz/lee/330
  • Reading list
  • I recommend the books by Tufte (2) and Cleveland
    (4)

21
General comment on resources
  • Getting books out of the library can be
    frustrating if you all do it at once
  • I recommend the use of the photocopier either in
    the City stats office (see Sharon Walker) or at
    Tamaki (see me)
  • Liaise with your classmates for making several
    copies of the same resource, use bulletin boards,
    emails (I will post a list of class addresses)
  • The City Stats office has copies of Venables and
    Ripley available for photocopying
  • I have some books that are also available from me
    for photocopying
  • There is good stuff available over the net Try
    the Departmental pages or Google
Write a Comment
User Comments (0)
About PowerShow.com