0% found this document useful (0 votes)
88 views24 pages

Linear Algebra

The document contains an example of a 3x3 matrix game. The optimal row and column strategies and the value of the matrix game are found. The saddle point is (-1, 2) with a value of -1. Reducing the matrix yields the optimal mixed strategies of playing the first row with probability 1/3 and the second column with probability 2/3.

Uploaded by

Prakriti Bhatia
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
88 views24 pages

Linear Algebra

The document contains an example of a 3x3 matrix game. The optimal row and column strategies and the value of the matrix game are found. The saddle point is (-1, 2) with a value of -1. Reducing the matrix yields the optimal mixed strategies of playing the first row with probability 1/3 and the second column with probability 2/3.

Uploaded by

Prakriti Bhatia
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 24

-alee Aad neor hlgebra

matiu A: 3 3
-3 -5 -3
3 3 1

3
-9 -5-A -3
3

-5-A 3-3 -3 3 3 5- = 0
3 3
3 3
(l)-5+5A e-A++ 3] -3[-3+Àtg + 3[-3 -(-15-34)]=0
(I-A)-5 +5A -A+A+) -3l-3+ 3a+31 +3[-3 +15+34J
((-[4 +4A+] -3[3 A46+3[6 +3¤Í =0
-to

for del -1 3 3
-3 -C -s
-3 -521 -3
3
-3 -6 -3 3 3
|-1 L3 3 3
+1 +2
2 l
3 3
33
3 3
0 3 -3J
2
V
V=
t
3 3
3 3
3 -3
-3
-3 -3
-3 -5+2 3
3 3 3

3 3 3
-3 -3 -3
Ry Ryt 3R
3 3 R:ky-3R, LO

magni tudu +*ó


Diagenal matri

-I

A PDSp

33 33
o -32 -33 -65 -33
33 33
a). Comashue a sachaldeangasiton g
C), Mek, achargs Opnohlil hat trons hu quodabc orm in pot b
néo aqutdnahc lom ait n9 CoS6- Peduct errm,paat b)Suludto te
tu
raicJormin hawum valu is
Con_hoat lulSod ioad ihd a urit wcl a at shich this
attinad.
). Alao, ynd ty mauinun xle 0. thun, Jinl a unt mcor
nce atot hich thi
Coscucs lell=1 and
maximum alu s atinud
>-4-A) - 4=0
4-d

-4-=o ’ - 5 =0 ’ACA-5) = o
M=5
tor dq=o, JA-Or| 4

for 45, IA-5I| -

JNormalyng the wcbes

L%
apucral decongoaikon

Areater than equal to guno, it is a aumi -dyirate


pooiti quadrahe orm.
2

Change in vartabl , Qlu) u 4uu, t tu,2


s

Subshhu hng U, à in he quad form


2 2
+

+ (-y -12 t l6)g2


(25 y*)= 59
as, thu eiyaton wth -bedat value
D. Qine tconstrants to lull 1,
maximum valu tu
5. quadratic 16

ho unt wcto a6s8ciaid asth eigurevalua 54

uh unit wctor
3

Ans - frd ATA.


A -3
-3-2-(2
-3-12 -12

9
D.in eignvalus oq ATA AA -A]|81-A

(8)(9)-slA -3A+-() -0

. Aind the eigenvedors o AA.


9AA -80t =

taa -81
Normaliged:atio

81

ormadigd lo

3 i0

Lo
Comput vectors U,U,Us
U Av, =

Qincs, Ay= 0, we naud to Suegincding


orthonormoú boais br Such ho
and y'Ug 0.

Normalgig x, we dotoain

9
2

oO

Ang 4=. And the covariance marix }or-hu data


from tha table qien
Mean t , = 120 +125 +125 t 135+ l4S, = 130
.

mean wtor, M1307


L65
aubtrocing rom hu caumns coe gt B=-io -5 -5 5 15
|-4 -5 -1 3

hu aamal covarian a manix is S= B8T

-4
|-4 -S -13
3 -5
-5
5
3
15 :5 25
etqenals Qru (o to deoimal placas
A =5526 t85 123.oa
Ag =-5\586 +l25 l.98.

dhu unit eigenclor coresponding toe A, o u |


O.436
for thu siza indux, &at.
wh Cundh
dariation yorm. hu coeiaht d huigh, Usp,in man
is l23. 0a,B cass thevaniance o< this index the datasut
tot
tHe si indax accounts tor variom
praca
ce is trSe l00 t25 = 12S
Havarianc the dadte ale o4o)
Wote dpasgl.
qs now exau hat
[=l,43.Ket us naus
om

Nbs, to Jird tu Scalans C and Cy uch that


bw veducd the
Qugmantid matrixhiuig hu pints a counns
O 5 5

Ca=- -

t indo eauahion.

8
3 Hencait is not
is
Os An induxed aat pints o ,V,.. vpS in K
not a9l
al ro
depundant thuu exist al numkys C,Ca,Cp not

Oheriaa,H at is ainly indapnct


L5
and n:|3 and A a,,a,,a, 3;Bb, b,, by 3. find a

B.Js the a
hy nplane H wth nor mal nthat saparatus AandAaad 8?
hypnglan pandlal to Hthat striciy spanata
Arta - ioling the eruahion 0.
Consid V=

ntajns thu normal wtor,and ten


ket a H, bu the orig in Hat cobalow
it can bu coriten as Shouon

So,t camn be coriten as (Ho9)- 3x, +2- 2z3


Ahygnplons is passing through & non-zho actors. oSo, th ale
shoon bow
d=G.
n and a, is ?
aling Ja)= a, e2)+t (-)-2(5)=-5
Ths dot praduck n and a, is
y(a2)n a, = 3(3) + I-()3) =4
The dot product oh and ag is ?
+ ()c) + (-2)Ce)
- (3)- )+6-I2
a,) ag =-3 = - 3 - G = -9,
The dot product on ond b is
- 3() +S- 2(-)=
The dot productn and by is : Cb)= hb = 3i)+ (-3) -a(-2)
product nand
Ad
BSepartes stricty 5J [: Hy
aparati stricty &Bnok A
ie H
upanáti stricty yis). -Hpnplane H
does
not :5 HHenoth
73 bythen B, and syanatu
AO ritty
hyprplont porallle), h 8, 4and <4eA)
equations abou the om
obawKd bu
Can
- And tha eptimal row and cslunn strodegies and thy value o thu
matrix game.

m in

3
ImoX

C
-1(- +at 5 =
Z= -) +tt2t = -1 + 3 t

t t2t +at}c2.

aaddle asint2

C= Lo
he matix is 5 -1 ’uduc it toae
4 a 3 matrix
2-3
min

3
hax 3
min max =2
Nog 9
poind vale
mal strate
optinal stratro
Plan l =
playe I Co,yp)
=
An
4 3
-3
qretn tham third vow
e d row s
|S column areta thom a column.
we reduce ou matrix. to a Qxa matrix,
Min

3
max
cahu tu
We haue veduud the prolalum to a point whu the
bust choiu is the only choiu
we re to solect hoiu al the ime.
shrattgy 9 we guta mctor wi'th exaty
., pu eq&lto 1and the est thu entresJ
One

equal to
:., Ophmal stratgs fow

column
Ogimd trag%

2
L- -3
L4 2 3J
valuu the a.
gone
&s- 3he Shady -an grass sud Componu alnds to typs sud
mistuts ungreen d Quckren.ach bad od Engreeh
pound d tach bag Quicegn contans
Qpoun ds d pound
laoo pounds
Jocu au 800 pounda o ry ased, onc 450 pounds bTogras
auaWabl to pt pt ito
te s mixhurs oThe cómpny maku
Quicarn that it producs..s u the mathunati cal
prolem that deter mind tha bags t each mìxtwe
mi
no.
t
that shoduy Lans shauld mak in dn t8t maximige ts pr

Sact tos 3z, +2x Si200

and c 2 o
First,uwe cupress echintauadity aoas an equation,using slack
Sinca,thr autheu in equaliies, ewe imroduca thet slackevoriabe,

M
3

&O0

a3 00
with thu botBorn ntry U50
r i,a
nagatiu obyciu Junchionas-2,-3 abn rom the
&-2, -32)
a.

"aina, tau ane no nuqahive numksu on thu right-most camn, we cam no


otoo
row -rezlatmunt on the lo wing tablaou.
3
800

-a -3 0
We perJorm roo-replacu mantCand pivatng) unlil all te entries in tu botom
row au non-nqate. The most agakve in ta bottom row is -3 and so
thu pivot is in the scond colum. Next, take tu quctients te
right- mast to u pivot cdumn . The aoat
Loi indicatt hu pivot rou
the poifuqucierts
IQ00a=600

450+ 450
Uo this cad,, thu daast
Jeast positu Quotient is in tu suond rouo. Thuss,
thu pivot is in the sucond row Scond column,
thu slecd pi vot is in thu
aueond rou suconcl column Weporm
row-repuacmt to the oth rous in a
Ssco nd celun ns entries coill b eqal to 0. Dividu the Sicond row
entries by 2so the vot coi b equal to 1?
3
400
450

-R2 +Ra> Rg
3Ra+ Ru4 ’Ry
to et the noys tabloau
400

L% 0 120
ina, tho 15 a hogai vee entry in the
sele a nouo pivo4t. bottom row, w
mostmgatie entry inin Hhu
tn t Jrst the botorm
botom row is
Ast column Net, tak tha quotierts andtthu Go the piyot iS
th pivot column. Thu tha right mst to
thu right-
pivot rous ot the posi
Aost posi htiw Quoñerts
quotients cooil indicat te

So this
Thusb, the pivot is in th positiu quotient is in t third row.
third row irst cocolum n,
he aaledd pivot is in
row replacmnt to tutte othen
third rorous dirst calumn. We perdorm
vos tn a coay thot han irst
row entries
entries by o tha vot oil bi be eaua to A
a
400

|(200
Agplg he Aolloing ro roplatmeds
-oo
to t the nw tobaus

350

Sincs, tha is no mor nagotive ernri es in tha btom row


twe haw ariud at the inaltaalau
T inal tablacus tho Variasls associctd Lsth each coumn
M

-| 350

1250
Qbbign aa val
We asig to yaniaklei the coumn it re prs
has etacty on 1 and th otho entries ue 0. Th
posititon ok the entryl indicat the vale
Xo
o
4 350
loo
1
the Soln is?
y=350
Ond
Ubad to
ohich us lad to ottain
ottain th the c ctive
maximum
1250 , ha au the irequaties, thuw, M=3.Sinc Hhthe amden
humbur t positi a; does not exud 9 that
hu soe" abovt is a basic Jacai bl, s J
thu, max prt $125o s attaind ohen l00 product
Ais producd'andl ssld, and hon 250 groduct BBis
þraducid seld.
Bhe max it ia50 uà ctainad caoon l0o product A
+
is praducad and sald, and cohen 350 prodc B
prod1c Bis producad
Cnd Sd.
M

Q. Detemina chich yariable shauld bu broughk into the soen


has do dc ter muiu which vaniabath shoutd be brsught into ta
/n the
t haue tu Aind most the caane
last row. Here -6 us the most Ngoi ic the ivst
Colum n L nudo to he broug ht into the Sol"
D. We alady heous faom (a) Hot a, à to bu cbrought ito the
chuck 20
min S 9 b2 3- min Slo,8 =8

nono the elements ohou kohw Hhe


pivot e as
M
3
-| o4
Rek-ak,
la
-G 5 3 |8

the next tobleou,


M

3 48

the
Rnow that
correspondunig to pantb)
into the Sol.
Ans we
. Whn x chck the cofuenks in at row hen u a
gatie no. - so the Uuunt scluthon is
So
Vnot tmal as 2 s prknt, Thrytoty nud to
Ainorm dn Hur iterakion the sinplex.
he
inegalies and b=ia
9

fhs t matrix inequality AzAz sbb yilda hu Jollousng syatem inequaltier:


ZI*,gualiy
n condihcn x0. laces hu holytepe in the
One wrtex is (o, o). he ,-inti cepts 3 hens.
(atn tao) any 1a 9 andG
amd6 so (6, o) ss a vwtix he ag-interapts
h 3 ins (uton e, =0) anu 4, 4o5 cn d12 so (0,4) s a verter.
the interauchion ca) and(b) i at Pa = (3,3)
Testing a in ) gives aca) ++ 1(3)=<j2 So Pab is in .
The inter tion o (6) and (c) is ot
Phe in a) ives I(5)+ 3(a) -|) i2 (5,a).Testin%
Pe so pbe is in pP.
Tu ioter a chon
tion (a) and c) is at Pac = (4.3,a..
Testing Pae in () qives i(4.8) +a(2.4) eBo6.
0, pae is not in P.
finally te ie wtices Cotrema poinds) e thu Polytepe
ae Co,o),6,o),(5, 2) (3,3and (o, 4).Thes points Jom
the minimal vugraatation o
Gsaically shaoing iti
1

12
18
(20), (o, 4)
3)
(3,o) (0,4)
(0, 1)
Asc <b 9

Pis a poins in R.
Lx 2

26
2
12

12

=3x=3 Pr =(3,3) PiG P.


Put t 3
(o,o) e f.

22

Pe=s,2
utin O
52 =4.8

(4.8 2.4)
Rt in
*+2
Yo8+4.8 9.6

minimal duprntatton,

You might also like

pFad - Phonifier reborn

Pfad - The Proxy pFad of © 2024 Garber Painting. All rights reserved.

Note: This service is not intended for secure transactions such as banking, social media, email, or purchasing. Use at your own risk. We assume no liability whatsoever for broken pages.


Alternative Proxies:

Alternative Proxy

pFad Proxy

pFad v3 Proxy

pFad v4 Proxy