Bayesian Variable Selection
Ho Chapter 9, Mixtures of gPriors Liang et al JASA
March 10, 2017
US Air Example
firms
popn
wind
precip
rain
SO2
Corr:
0.19
Corr:
0.0627
Corr:
0.35
Corr:
0.386
Corr:
0.43
Corr:
0.434
temp
Corr:
0.955
Corr:
0.238
Corr:
0.0324
C
Bayesian Estimation in Linear Models
STA521 Linear Models Duke University
Merlise Clyde
February 19, 2017
Bayesian Estimation
Model Y = X + with N(0n , 2 In ) is equivalent to
Y N(X, In /)
= 1/ 2 is the precision.
In the Bayesian paradigm describe uncert
Generalized Ridge & Lasso Regression
Readings ISLR 6, Casella & Park
STA 521 Duke University
Merlise Clyde
March 20, 2017
Model
Model: Y = 10 + X + X is centered and scaled predictors
(Classical) Ridge Regression controls how large coecients
may grow
X)T
Use svm() in R to fit the default svm to the last
column of the sonar training data at
sonar_train.csv
Compute the misclassification error on the
training data and also on the test data at
sonar_test.csv
Solution:
install.packages("e1071")
library(e1071)
Stochastic MultiArmed bandit Algorithm
Stefano Traca and Cynthia Rudin
1
Problem setup
The name multiarmed bandit comes from the name of a gambling machine. You can choose one of the arms (levers) of the
machine at each round, and get a reward based on
Multi

armed
Bandits
MAB
gambling
a

*
machine
T T
T
/
.
(
#
at

t
time
,
play
I
when
neo to explore
exploit good
and
j
arm
oew_ardyx.tt
with
(
suffer
A;
expected
=
,*
of
my
estimate
current
I;
of
,
:
"
=pt"
X ; G)
FEI

best
t
j before
played
Total r
9/20/16
Boosting
Cynthia Rudin
Machine Learning Course, Duke
Boos+ng Mo+va+on
Question of Kearns: Can you turn a weak learning
algorithm (that is barely better than random guessing) into a
strong learning algorithm (whose error rate is arbitrarily
clos
9/8/16
Cross Validation
Cynthia Rudin
Machine Learning Course, Duke
CrossValidation
I tried ridge regression and I tried least square regression,
how do I know which one performed better?
1
9/8/16
CrossValidation
Cross Validation (CV) is the most p
Stats 202 Practice Problems
August 7, 2015
1. We fit a linear regression model Y = 0 +1 X1 +. . .+p Xp to some data. Suppose we change
the units of the predictors Xi , to obtain a new set of predictors Zi = cXi . Then, we fit the
same data to the model: Y
Causolynfeoencl
Is
for
useless
ML
dad
Possibly
"
"
male
C)
68
fcx )
predict
of ( gender
.
#
to t
.
stops
If
f ( )
=
2
*
smoking
what
?
Away
,
it
does
ftp.warfa.in ]
#
.
4 '% #
24
+
]
etc
,
)
.
,
+
]
drugs
Musty
age ,
,
heart future

Correlation
happened
Module 2: Introduction to Decision Theory
Rebecca C. Steorts
Exercise
Suppose X  Bin(n, ) and  Beta(a, b). Derive the posterior
distribution of . Now derive the marginal distribution p(x ). How
does this differ from the BernoulliBeta example? Is one a
Module 1: Introduction to Bayesian Statistics,
Part I
Rebecca C. Steorts
Agenda
I
I
I
I
I
I
I
I
I
Motivations
Traditional inference
Bayesian inference
Bernoulli, Beta
Connection to the Binomial distribution
Posterior of BetaBernoulli
Example with 2012 el
Lab 4: Bayesian Statistics in R  STA 360/602
Rebecca C. Steorts
1
Agenda
Do a teachers expectations influence student achievement? In a famous study,
Rosenthal and Jacobson (1968) performed an experiment in a California elementary school to try to answer
STA 360/602: Bayesian Methods and Modern Statistics
Duke University, Spring 2017
Instructor : Rebecca C. Steorts, Assistant Professor, Dept of Statistical Science, [email protected]
Course Time: T/Th: 11:45 am 1:00 pm
Steorts Office Hours: Tu: 1:002:00
Lab 3: Introduction to Decision Theory in R STA 360/602
Rebecca C. Steorts
1
Agenda
In class, you saw the resource allocation example. We will now go through how
to reproduce parts of the lecture using R in Tasks 12 and Tasks 35 should be
completed in yo
Lab 2: Bayesian Statistics in R  STA 360/602
Abbas Zaidi and Rebecca C. Steorts
1
Agenda
In class, you saw the BinomialBeta model. We will now use this to solve a
very real problem! Suppose I wish to determine whether the probability that
a worker will
STA360/602 Exam I, Spring 2017
Instructions
Write your name, NetID, and signature below.
If you need extra space for any problem, continue on the back of the page.
Circle if you are in STA 360 or STA 602.
Community Standard
To uphold the Duke Community
Lab 1: Introduction to Programming in R
STA 360/601
Abbas Zaidi
1
Agenda
1. Installing RStudio.
2. Markdown and why you should know it.
3. Initializing variables: storing, some variable types, length of vectors, summary statistics, printing and seeding
4
Business Law  Honors
Objective 1.01
Ethics, Sources of Law & Legal Systems
1
Trolley Car / Doctor
Runaway Trolley Car You are the driver
Can not stop the trolley  5 workers on track will die
If you turn onto alternate track, 1 worker on that track wi
Business Law  Honors
Objective 1.01
Ethics, Sources of Law & Legal Systems
1
Trolley Car / Doctor
Runaway Trolley Car You are the driver
Can not stop the trolley  5 workers on track will die
If you turn onto alternate track, 1 worker on that track wi
MATTHEW BOEZEMAN
BB30 Business Law 3.03 Summer 2013
Environmenta
l & Energy
Law
Environmental Law
BB30 Business Law 3.02 Summer 2013
Place a clear, noncopyrighted
picture that represents environmental
law on this slide.
Describe the EPA:
When was the EPA
BL_1.03 Notes_Activity
Name: Matthew Boezeman
Date: 2/17/2017
Notes for 1.03
Understand Criminal and Civil Laws
An act that is punishable offense against society
Criminal Act
Elements of
Criminal Acts
Duty
Whether you were aware of your duty to do or not
5.01Activity 1 TrueFalse for Terms
_1. A valid premarital agreement may be oral
_2. The marriage contract comes into existence when the couple becomes
engaged
_3. The primary duty that arises from a marriage contract is the duty financial
support.
_4. A
02.02P3 Business Law HONORS
Unit A:
Competency:
Objective 2.02:
Business Law
Understand Contract Law
Understand terminating, transferring & breaching contracts
Contract Law Case Studies
Instructions:
Use the information you have learned in Objective 2.01
Chapter 9 Designing
Classes
Chapter Goals
Learn to identify side effects
Know when to use method preconditions
and postconditions
Go in depth a bit more on static (aka class)
methods and fields
Design programs the objectoriented way
9.3 Immutable Cla
5.01 Activity 2 Completion
1. An engaged couple has a _ marriage contract.
2. The legal term for the age at which marriage is permitted is the age of
_.
3. The personal relationship between two persons who are united as husband and
wife is called_.
4. Mar
3.01 Activity 5B
Employment Discrimination Lawsuit
OPTION #2: Mock Trial
Students can select a recent class action lawsuit involving employment discrimination
and answer the questions from the previous page. Students are present the information
in the for
5.02 Activites_Key_Terms
Attorneyinfactthe person receiving the power of attorney
Beneficiaryis someone who receives property by a will
Bequest or legacya gift of personal property made by a will
Capacitymeans that the person that is making a will i