StatC180/C236
Introduction
to
Bayesian
Statistics
Course syllabus
Instructor: Juana Sanchez
Office hours: Wednesday
4-5
pm
and
Friday,
4-5
pm
and
also
by
appointment
Metropolis-Hoff
Chapter 10-finish
Prior distributions
Hwk 5 due
Thursday
June
10
10 min max
presentation of your
project in power point
MS 5128.
(b) Turn in final written
version of your
project. No amendments
will be allowed after
that. No late projects
accepted under any
circumstances.
BAYES
SITES
ISBA
SBSS (ASA)
WINBUGS Manuals
(in Intro to WB)
Required textbook
(PDH): free access to chapters from our library.You may download
Note percentage points for each part; they can not be made up; all
R code
and output must battached in appendix, with comments and sections)
Friday, April 16: (1)
Identify
several
articles
and
books
in
the
literature
that
have
studied
the
question
that you proposed. Start typing a section in
your project where you write: the question you are interested in
(revised after getting my comments) and what this literature has found
to answer this question. Attach your articles to the the typed section
and turn in the typed section and the articles. You will need all this
information later to assess your prior distributions (3 pts)
Friday, April 23: Your data must be ready. Add a new
typed
section of your paper where you describe the sources of
the data and describe the variables from the data set that you will
use, whether they are outcomes or explanatory variables, sample
size, type of data and code and units (categorical, numerical) and how
it was obtained (i.e.,
experimental,
observational,
etc),whether
there
are
missing
values
or
not.
Print
a
few
lines
(no
more
than
5)
of
the
data
that
you
will
use
and
put
in
an
appendix.
Address issues such as wether the data are
in R format already. Turn in the updated paper, with the
revised section
that
your
turned
in
on
April
6
and
the
new
section
describing
the
data.
Either
send me the data or
stop by my office with your computer to show it to me. It is very
important that you talk only about the data you will use, even if this
is just part of a bigger data set. (3 points)
Friday, May 14: Decision
on
Likelihood
model
and
prior
to
be
used.
Write
a
section
justifying
the
use
of
these
models.
Unknowns
to
do
inference
on.
Assessment
of
priors
on
the
parameters.
Turn
in
all
the things done up to now plus
new section typed. Even though you may have hinted earlier what model
you will use etc.. you still need to do this step. (3 points)
Friday,
May 28: Inference and prediction. Preliminary posterior
and
predictive results must be available (ok if done with simpler models
than what you wanted to do). A typed section of your paper
will contain the results and discussion of the results. Turn in updated
paper with revisions of prior phases. Explanation of
computational methods used must be included in the narrative. (5
points). Do not turn in just one section. Turn in the whole paper. The
model used must be written in math form ( theta ~ binom(n, p_i ) ),
etc...
June 10: Conclusions section,
all revisions must have been completed. Turn in the final paper.
Presentation in front of the class. 10 minutes. The paper turned in is
final. See handout with instructions given to you in class.
HOMEWORK
Homework 1 (Due
4/9) Hwk
1
key
Hwk 2 key
Homework 3
Due May 3
Odds/risk
Hwk 3 key
Homework
4
Due May 19
Hwk 4 key
(problem 2
still to be typed).
Homework 5
Due June 4
HWK HINTS
R
CORNER
sleep1.r
(1.4)