Scribd
Upload
143 views

Pattern Recognition Problem Set 1 PDF

This document contains instructions for Problem Set 1 of a pattern recognition course. It includes 7 problems related to topics in probability, high-dimensional spaces, Bayesian inference, covariance matrices, distribution properties, and a simulation of the Monty Hall problem. Students are asked to analyze concepts, prove statements, plot graphs, write programs, and simulate scenarios. They may use Python or MATLAB for programming assignments.

Uploaded by

lmaraujo67
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
143 views

Pattern Recognition Problem Set 1 PDF

This document contains instructions for Problem Set 1 of a pattern recognition course. It includes 7 problems related to topics in probability, high-dimensional spaces, Bayesian inference, covariance matrices, distribution properties, and a simulation of the Monty Hall problem. Students are asked to analyze concepts, prove statements, plot graphs, write programs, and simulate scenarios. They may use Python or MATLAB for programming assignments.

Uploaded by

lmaraujo67
Copyright
© © All Rights Reserved
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 3

ProblemSet1

MAS 622J/1.126J: Pattern Recognition and Analysis


Due Monday, 18 September 2006
[Note: Allinstructionstoplotdataorwriteaprogramshouldbecarriedoutus-
ingeitherPythonaccompaniedbythematplotlibpackageorMatlab. Feelfree
touseeitherorboth,butinordertomaintainareasonablelevelofconsistency
and simplicity we ask that you do not use other software tools.]
Problem 1: Why?
a. Describe an application of pattern recognition related to your research.
What are the features? What is the decision to be made? Speculate on
how one might solve the problem. Limit your answer to a page.
b. Inthesameway,describeanapplicationofpatternrecognitionyouwould
be interested in pursuing for fun in your life outside of work.
X
Problem 2: Probability Warm-Up
Let X and Y be random variables. Let
X
E[X] denote the expected value
ofX and
2
E[(X
X
)
2
]denotethevarianceofX. Useexcruciatingdetail
to answer the following:
a. Show E[X+Y]=E[X]+E[Y].
X
b. Show
2
=E[X
2
]
2
.
X
2
are uncorrelated, then
Z
c. Show that independent implies uncorrelated.
d. Show that uncorrelated does not imply independent.
e. Let Z = X +Y. Show that if X and Y =
.
2 2
+
Y X
f. Let X
1
and X
2
be independent and identically distributed continuous
random variables. Can Pr[X
1
X
2
] be calculated? If so, nd its value.
If not, explain.
g. LetX
1
andX
2
beindependentandidenticallydistributeddiscreterandom
variables. Can Pr[X
1
X
2
] be calculated? If so, nd its value. If not,
explain.
1



Problem 3: High-Dimensional Probability
LetX= (X
1
, X
2
, . . . , X
n
)bearandomvector,wherethe{X
i
}areindependent
and identically distributed (i.i.d.) continuous random variables with a uniform
probability density function between 0 and 1:
1,for 0x
i
1
p(x
i
) =
0,otherwise
Each value x of the random vector X can be considered as a point in a n-
dimensionalhypercube. SincetheprobabilitydensityfunctionofXisuniform,
volume in this n-dimensional space corresponds directly to probability. Find
anexpressionforthepercentageofan-dimensionalhypercubesvolumelocated
within of the hypercubes surface. Plot this percentage as a function of n for
1n100000and= 0.0001. Whatdoyourndingsabouthigh-dimensional
hypercubes tell you about random variables?
Problem 4: Teatime with Gauss and Bayes
1 2
2
2
2
Let p(x,y) =
2
e

(y)
2
+
(xy)
2
.
a. Find p(x), p(y), p(x|y), and p(y|x). In addition, give a brief description
of each of these distributions.
b. Let=0,=40,and =3. Plotp(y)andp(y|x=13.7)forareasonable
range of y. What is the dierence between these two distributions?
Problem 5: Covariance Matrix
Let
X
=
64 25
.
25 64
a. Verify that
X
is a valid covariance matrix.
b. Findtheeigenvaluesandeigenvectorsof
X
byhand. Showallyourwork.
c. Writeaprogramtondandverifytheeigenvaluesandeigenvectorsof
X
.
d. We provide 200 data points sampled from the distribution N(0,
X
).
Download the dataset from the course website and plot the data points.
Project the data onto the covariance matrix eigenvectors and plot the
transformed data. What is the dierence between the two plots?
2

Problem 6: Distribution Linearity


Let X
1
and X
2
be i.i.d. according to
1,for 0x
i
1
p(x
i
) =
0,otherwise
for i= 1,2
Let Y =X
1
+X
2
.
a. Find an expression for p(y). Plot p(y) for some reasonable range of y.
b. Find an expression for p(x
1
|y). Plot p(x
1
|y) as a function of x
1
with y
treated as a known parameter for some reasonable value of y and some
reasonable range of x
1
.
c. Repeat the parts above, this time letting X
1
and X
2
be i.i.d. according
to N(0,1).
d. What was the point of this problem? Hint: check out the title.
Problem 7: Monty Hall
Togetcreditforthisproblem,youmustnotonlywriteyourowncorrectsolution,
butalsowriteacomputersimulation(ineitherMatlaborPython)oftheprocess
of playing this game:
Suppose I hide the ring of power in one of three identical boxes while you
werent looking. The other two boxes remain empty. After hiding he ring of
power, I ask you to guess which box its in. I know which box its in and, after
youve made your guess, I deliberately open the lid of an empty box, which is
oneofthetwoboxesyoudidnotchoose. Thus,theringofpoweriseitherinthe
box you chose or the remaining closed box you did not choose. Once you have
madeyourinitialchoiceandIverevealedtoyouanemptybox,Ithengiveyou
the opportunity to change your mind you can either stick with your original
choice, or choose the unopened box. You get to keep the contents of whichever
box you nally decide upon.
What choice should you make in order to maximize your chances of re-
ceiving the ring of power? Explain your answer.
Writeasimulation. Therearetwochoicesinthisgameforthecontestant
inthisgame: (1)choiceofbox,(2)choiceofwhetherornottoswitch. In
your simulation, rst let the host choose a random box to place the ring
of power. Show a trace of your programs output for a single game play,
as well as a cumulative probability of winning for 1000 rounds of the two
policies (1) to choose a random box and then switch and (2) to choose a
random box and not switch.
3

You might also like