ie256p3

8/9/2019 ie256p3

1/54

DECISION MAKING

ObjectivesRelation to Planning

Types of Decisions

Discssion on Mo!eling

Types of Decision Ma"ing

Decision Ma"ing #n!e$ Ce$tainty% &inea$ P$og$a''ing

Decision Ma"ing #n!e$ Ris"% e(pecte! vale) !ecisiont$ees) *eing t+eo$y) an! si'lation

Decision Ma"ing #n!e$ #nce$tainty% Ga'e T+eo$y

Integ$ate! Data ,ases) MIS) DSS) an! E(pe$t Syste's

[email protected]

8/9/2019 ie256p3

2/54

DECISION MAKING

Relation to PlanningDecision Ma"ing% P$ocess of 'a"ing a conscios c+oice bet-een

. o$ 'o$e alte$natives p$o!cing 'ost !esi$able conse*ences

/benefits0 $elative to n-ante! conse*ences /costs01

Decision Ma"ing is essential pa$t of Planning

Planning% Decision in a!vance -+at to !o) +o- to !o) -+en to

!o an! -+o is to !o it1

Re*i$e! also in

Designing an! Staffing an! O$gani2ation Developing Met+o!s of Motivating Sbo$!inates

I!entifying Co$$ective Actions on Cont$ol P$ocess

[email protected]

8/9/2019 ie256p3

3/54

Occasions fo$ DecisionOccasions a$e in 3 !istinct fiel!s%

4$o' At+o$itative Co''nications f$o' spe$io$s

4$o' Cases Refe$$e! fo$ Decision by Sbo$!inates

4$o' Cases O$iginating in t+e Initiative of t+e E(ective Most i'po$tant test of e(ective

DECISION MAKING

[email protected]

8/9/2019 ie256p3

4/54

Types of DecisionsRotine Decisions /e1g1 pay$oll p$ocessing) paying spplie$s etc0 Rec$ f$e*ently

Involve Stan!a$! Decision P$oce!$es

5as a Mini'' of #nce$taintySt$ct$e! Sitations

Non$otine Decisions

#nst$ct$e! an! Novel Sitations

Non$ec$$ing Nat$e 5ig+ &evel of #nce$tainty

DECISION MAKING

[email protected]

8/9/2019 ie256p3

5/54

Objective ve$ss ,on!e! RationalityA Decision is objectively $ational if it is t+e co$$ect be+avio$ fo$'a(i'i2ing given vales in a given sitation1

Rationality $e*i$es

61 A co'plete "no-le!ge an! anticipations of conse*ences

afte$ a c+oice.1 I'agination since Conse*ences lie in ft$e

.1 A c+oice a'ong all possible alte$natives1

7e can only tal" abot bon!e! $ationality

DECISION MAKING

[email protected]

8/9/2019 ie256p3

6/54

Objective ve$ss ,on!e! Rationality

Objective Rationality looks for thebest solution whereas BoundedRationality looks for the goodenough solution.

DECISION MAKING

[email protected]

8/9/2019 ie256p3

7/54

Manage'ent Science C+a$acte$istics61 A Syste' 8ie- of P$oble'.1 Tea' App$oac+

31 E'p+asis on t+e #se of 4o$'al Mat+e'atical Mo!els an!

Statistical an! 9antitative Met+o!s

Mo!els:

DECISION MAKING

[email protected]

8/9/2019 ie256p3

8/54

Mo!els an! T+ei$ Analysis

Mo!el% Abst$action an! Si'plification of Reality

/Designe! to incl!e Essential 4eat$es0

DECISION MAKING

Si'plest Mo!elnet income = revenue expenses -taxes

[email protected]

8/9/2019 ie256p3

9/54

DECISION MAKING

; Steps of Mo!eling

Real 7o$l! Si'late! /Mo!el0 7o$l!61 4o$'late P$oble'

/Define objectives) va$iables an! const$aints0

.1 Const$ct a Mo!el /si'ple bt

$ealistic $ep$esentation of syste'0

31 Test t+e Mo!el

8/9/2019 ie256p3

10/54

DECISION MAKING

Catego$ies of Decision Ma"ing

Decision Ma"ing #n!e$ Ce$tainty% &inea$ P$og$a''ingDecision Ma"ing #n!e$ Ris"% e(pecte! vale) !ecision

t$ees) *eing t+eo$y) an! si'lation

Decision Ma"ing #n!e$ #nce$tainty% Ga'e T+eo$y

[email protected]

8/9/2019 ie256p3

11/54

DECISION MAKING

Payoff /,enefit0 Table > Decision Mat$i(

N1 N2 Nj Nn

P1 P2 Pj Pn

A1 O11 O12 O1j O1n

A2 O21 O22 O2j O2n

. .

Ai Oi1 Oi2 Oij Oin. .

Am Om1 Om2 Omj Omn

Alte$nativeState of Nat$e ? P$obability

Otco'e

S' of nvales ofpj'st be 6 [email protected]

8/9/2019 ie256p3

12/54

DECISION MAKING

Decision Ma"ing #n!e$ Ce$tainty

I'plies t+at -e a$e ce$tainof t+e ft$e state of nat$e

/o$ ass'e -e a$e0

T+is 'eans%> the probability o pjo uture Njis 1 an! all

other utures have "ero probability.

[email protected]

8/9/2019 ie256p3

13/54

DECISION MAKING

Decision Ma"ing #n!e$ Ris"

T+is 'eans%> #ach Njhas a $no%n &or assume!' probability o pjan!

there may not be one state that results best outcome.

[email protected]

8/9/2019 ie256p3

14/54

DECISION MAKING

Decision Ma"ing #n!e$ #nce$tainty

T+is 'eans%> Probabilities pjo uture states are un$no%n.

[email protected]

8/9/2019 ie256p3

15/54

DECISION MAKING

Decision Ma"ing #n!e$ Ce$tainty

&inea$ P$og$a''ing

A !esi$e! benefit /p$ofit0 e(p$esse! as a a 'at+e'atical fnction of

seve$al va$iables1 Soltion is to fin! in!epen!ent va$iables giving

t+e 'a(i'' benefit sbject to ce$tain li'its /to const$aints01

[email protected]

8/9/2019 ie256p3

16/54

DECISION MAKING

Decision Ma"ing #n!e$ Ce$tainty &inea$ P$og$a''ing

E(a'pleA facto$y is p$o!cing t-o p$o!cts /(an! )01

@6 p$ofit pe$ nit of p$o!ct( an! @6= pe$ nit of p$o!ct )1

7+at is t+e p$o!ction level ofxnits of p$o!ct(an!ynits of

p$o!ct )t+at 'a(i'i2es t+e p$ofitP:

Ma(i'i2ePB6x6=y

[email protected]

8/9/2019 ie256p3

17/54

DECISION MAKING

Decision Ma"ing #n!e$ Ce$tainty &inea$ P$og$a''ing

Ma(i'i2ePB6x6=y

6 . 3 = ; F

6

=

3

.

;

#nits of p$o!ctx

#nitsofp$o!cty

4o$ e(a'ple fo$ nits of

-e get P of /6H0

An! f$o' PB6=y

B6=y yB;

PB

Isop$ofit line

@ by selling nits of o$ ; nits of

@. by selling . nits of o$ ==13 nits of

@3; by selling 3; nits of o$ .; nits of

PBD.A

PB3;

A

PBEAA

[email protected]

&i P i

8/9/2019 ie256p3

18/54

6 . 3 = ; F

6

=

3.

;

#nits of p$o!ctx

#nitsofp$o!cty

*onstraint 1

+x2y 6.

PBD.A

*onstraint 2

x , 2y F

Ma(i'' P$ofit point

-it+in const$aints

DECISION MAKING Decision Ma"ing #n!e$ Ce$tainty &inea$ P$og$a''ing

o$ p$o!ction /an! p$ofit0 is sbject to $eso$ce li'itations) o$ constraints.

o e'ploy ; -o$"e$s /3 'ac+inists an! . asse'ble$s0) eac+ -o$"s only = +o$s a -ee"1

*ON/A0N

Pro!uct ( reuires + hours o machinin an! 1 hour assembly per unit

Pro!uct ) reuires 2 hours o machinin an! 2 hours o assembly per unit

*onstraints expresse! mathematically

1. +x2y 6. &hours machinin time'2. x , 2y F &hours assembly time'

P=10x+14y

=20*10+30*14

=620

[email protected]

&i P i

8/9/2019 ie256p3

19/54


Dr.B.G.Cetinercetiner itu.edu.tr

Co'pany 'a"es t-o !es"s%

Type

Material Usage

Profit7oo! Metal Plastic

Re! 6 = 6; 66;

,le . 6 6 J

Available Ra- Mate$ial%

7oo! .Metal 6.FPlastic ..

Steps:

Determine decision variables! "hey are red and blue desks.

Determine objective function! #$%%&.'%()*.'+

Determine constraint functions and ,lot them!

..66;

6.F6=

..6

.

6

.6

.6

.6

+

+

+

x

x

xx

xx

xx

Ma(i'i2e p$ofit

C S O A G &i P i

8/9/2019 ie256p3

20/54


Dr.B.G.Cetinercetiner itu.edu.tr

G.

G6

6;(6C 6A (.B ..A

=(6C 6D (.B 6.F

6A(6C .A (.B .AA

-easible

Region

O,timum X%

O,timumX+

Opti'' points can

be fon! g$ap+ically

#$%%&.'%()*.'+#$%%&.%+()*.#$%/* 0a1imum #rofit

23O#RO-2" 4256

6.

=

DECISION MAKING D i i M "i # ! C i &i P i

8/9/2019 ie256p3

21/54


*O3P4#/ O540ON

0n reality6 %e have more than 2 variables &!imensions' not li$e machinin an!assemblin only.

*omputer solution calle!Simplex metho! has been !evelope! to be use! %ith

many variables.

7or example6 an A8 mo!el o current an! uture telephone !eman! has ot

92::: variables ta$in 9 to ; computer hours or sinle run.

7or example6 one mo!el o the !omestic lon-!istance net%or$ has

base! on ?projective eometry@ that is : to 1:: times aster.

> ?he tartlin Biscovery Cell 5abs ept in ha!o%s@6 Cusiness Dee$6

eptember 216 1E

8/9/2019 ie256p3

22/54


GO3#DO/ 1

)ou operate a small %oo!en toy company ma$in t%o pro!uctsH alphabet bloc$san! %oo!en truc$s. )our proit is I+: per box o bloc$s an! I9: per box o truc$s.

Pro!ucin a box o bloc$s reuires one hour o %oo!%or$in an! t%o hours o

paintinJ pro!ucin a box o truc$s ta$es three hours o %oo!%or$in but only one

hour o paintin. )ou employ three %oo!%or$ers an! t%o painters6 each %or$in

9: hours a %ee$. Go% many boxes o bloc$s &C' an! truc$s &' shoul! you ma$eeach %ee$ to maximi"e proitK olve raphically &usin millimetric paper' as a

linear proram an! conirm analytically.

[email protected]

DECISION MAKING

8/9/2019 ie256p3

23/54

DECISION MAKING


T+e$e e(ist a n'be$ of possible ft$e states of Nat$eNj.

Eac+Nj+as a "no-n /o$ ass'e!0 p$obabilitypjof occ$$ing1

T+e$e 'ay not be one ft$e state t+at $eslts in t+e best otco'e fo$

all alte$nativesAi.

[email protected]

DECISION MAKING

8/9/2019 ie256p3

24/54

DECISION MAKING


*alculate #xpecte! Lalues i'olution

n

j=1

&pjOij'#i=*hoose the Alternative Aiivin

the hihest expecte! value

[email protected]

DECISION MAKING Decision Ma"ing #n!e$ Ris" *alculate #xpecte! Lalues i'

8/9/2019 ie256p3

25/54

n

j=1

&pjOij'#i=

DECISION MAKING Decision Ma"ing #n!e$ Ris"*alculate #xpecte! Lalues i'

#1=:.EEE>&-2::',:.::1>&-2::' #1=I-2::

#2=:.EEE>:,:.::1>&-1::6:::' #2=I-1::

E(a'ple of Decision Ma"ing #n!e$ Ris"

Not 7ire

in your house

P1 =:.EEE P2=:.::1

0nsure house I-2:: I-2::

Bo not 0nsure house : I-1::6:::

Alte$natives

7ire

in your house

State of Nat$e

P$obabilities

Doul! you insure your house or notK


8/9/2019 ie256p3

26/54


7ell D$illing E(a'ple>Decision Ma"ing #n!e$ Ris"

N1HBry Gole N2Hmall Dell N+HCi Dell

P1=:.F P2=:.+ P+=:.1

A1HDon;) @3) @J)3)

Alte$nativeState of Nat$e ? P$obability

A+H4a$' Ot @ @6.;) @6).;)

#xpecte!

Lalue

#1=:.F>:,:.+>:, :.1>:

@

#2=:.F>&-::6:::',:.+>&+::6:::', :.1>&E6+::6:::'

@.)

#+=:.F>:,:.+>&126:::', :.1>&162:6:::'

@6.)

@.)A2is t+e soltion if yo a$e -illing to $is" @;)


8/9/2019 ie256p3

27/54


Becision rees

Ins

$e

Don

8/9/2019 ie256p3

28/54

DECISION MAKING Decision Ma"ing #n!e$ Ris"

Mueuin &Daitin 5ine' heory

*lass o People

Or

ObjectsArrivals

Se$ve$s

:Ti'e bet-een

a$$ivals :Se$vice ti'e

$e*i$e! fo$

eac+ a$$ival

[email protected]

DECISION MAKING Decision Ma"ing #n!e$ Ris" Mueuin &Daitin 5ine' heory

8/9/2019 ie256p3

29/54

Arrivals


Se$ve$s

0!entiy optimum number

o servers nee!e! to re!uce

overall cost.

erve!

Dr.B.G.Cetiner


8/9/2019 ie256p3

30/54


Se$ve$s



overall cost.

Arrivals Arrivals Arrivals Arrivals

Mueue

Dr.B.G.Cetiner


8/9/2019 ie256p3

31/54


Se$ve$s



overall cost.

erve!Arrivals Arrivals Arrivals

Dr.B.G.Cetiner


8/9/2019 ie256p3

32/54


:Ti'e bet-een

a$$ivals :Se$vice ti'e

$e*i$e! fo$

eac+ a$$ival

Approximate! byProbability Bistribution


[email protected]


8/9/2019 ie256p3

33/54


ypical Daitin-5ine ituations


[email protected]

O$gani2ation Activity A$$ivals Se$ve$s

Airport 5an!in Airplanes /un%ay

Personnel Oice ob 0ntervie%s Applicants 0ntervie%ers

*ollee /eistration tu!ents /eistrars

*ourt ystem rials *ases u!es

Gospital 3e!ical ervice Patients /oomsBoctors

upermar$et *hec$out *ustomers *hec$out cler$soll bri!e a$in tolls Lehicles oll ta$ers

oolroom ool issue 3achinists oolroom cler$s

DECISION MAKING Decision Ma"ing #n!e$ Ris" imulation

8/9/2019 ie256p3

34/54


imulation

imulation

[email protected]

In case $eal>-o$l! syste' is too co'ple( to e(p$ess in si'ple e*ations1

Soltion is to const$ct a 'o!el t+at si'lates ope$ation of a $eal syste' by

'at+e'atically !esc$ibing be+avio$ of in!ivi!al inte$$elate! pa$ts1

Ti'e bet-een a$$ivals an! se$vices can be $ep$esente! by p$obability !ist$ibtions1

Bevelop a computer proram or one cycle o operation6an! /un it or many cycles.

DECISION MAKING Decision Ma"ing #n!e$ Ris" /is$ as Lariance

8/9/2019 ie256p3

35/54


/is$ as LarianceH example

/is$ as Lariance

[email protected]

P$oject P$oject P$obability Cas+ 4lo- P$obability Cas+ 4lo-

16 @3 16 @.

1. @3; 1.; @3

1= @= 13 @=

1. @=; 1.; @;

16 @; 16 @

#xpecte! *ash 7lo%s#&x'=:.1:&+:::',:.2:&+::',:.9:&9:::',:.2:&9::',:.1:&:::'

=I9:::#&y'=:.1:&2:::',:.2&+:::',:.+:&9:::',:.2&:::',:.1:&F:::'

=I9:::

Dhich one %oul! you chooseK

DECISION MAKING Decision Ma"ing #n!e$ Ris" /is$ as Lariance

8/9/2019 ie256p3

36/54


/is$ as LarianceH example

/is$ as Lariance

[email protected]

5oo$ at the variances or stan!ar! !eviations6An! choose the one %ith lo%est variance &or !eviation'

L&x'=:.1:&+:::-9:::'2,:.2:&+::-9:::'2,.,:.1:&:::-9:::'2

=+::6:::L&y'=:.1:&2:::-9:::'2,:.2&+:::-9:::'2,..,:.1:&F:::-9:::'2

=16+::6:::

&x'=I90/-o$st otco'e0

@6)=;)H

L1./J)3)30/6>1.0/>;)0

L1./6).;)30/6>1.0/0

@.;)

Optimist Pessimist

DECISION MAKING Becision 3a$in 4n!er 4ncertainty

8/9/2019 ie256p3

40/54


[email protected]

Becision 3a$in 4n!er 4ncertaintyH 3aximum /eret

N1HBry Gole N2Hmall Dell N+HCi Dell

De !o not $no% probabilities

A1HDon

8/9/2019 ie256p3

41/54


[email protected]

Becision 3a$in 4n!er 4ncertaintyH ame heory

7uture states o natures an! their probabilities are replace!

by

the !ecisions o competitor6 or their strateies.

ry to *atchtrateies o your competitor.


8/9/2019 ie256p3

42/54


[email protected]

Becision 3a$in 4n!er 4ncertaintyH ame heory

7or example6 OBB an! #L#N ame.

%o players lash one or t%o iners.

0 the total is 2 or 9 then #ven %ins60 it is + O!! %ins.

DECISION MAKING

8/9/2019 ie256p3

43/54

DECISION MAKING

[email protected]

0nterate! Bata Cases6 306 B an! #xpert ystems

DECISION MAKING

8/9/2019 ie256p3

44/54

[email protected]

T$ansaction P$ocessing Syste's /TPS0

Manage'ent Info$'ation Syste's /MIS0

Decision Sppo$t Syste's /DSS0

9 ypes o 0normation ystems

E(pe$t Syste's /ES0

DECISION MAKING

8/9/2019 ie256p3

45/54

[email protected]

ra!itional Approach an! Bata &4ser' Oriente! Approach

Pay$oll

Syste'

P$ojectManage'ent

Syste'

Ta(Data

Pe$sonnelData

P$ojectsData

Pe$sonnelData

ra!itional Approach

DECISION MAKING

8/9/2019 ie256p3

46/54

[email protected]

ra!itional Approach an! Bata &4ser' Oriente! Approach

Pay$oll

Syste'

P$ojectManage'ent

Syste'

Ta(Data

P$ojectsData

Pe$sonnelData

Batabase Approach

DECISION MAKING

8/9/2019 ie256p3

47/54

[email protected]

T$ansaction P$ocessing Syste's /TPS0 Ato'ate +an!ling of !ata abot bsiness activities

/t$ansactions0

Manage'ent Info$'ation Syste's /MIS0 Conve$ts $a- !ata f$o' t$ansaction p$ocessing syste'

into 'eaningfl fo$'

Decision Sppo$t Syste's /DSS0 Designe! to +elp !ecision 'a"e$s

P$ovi!es inte$active envi$on'ent fo$ !ecision 'a"ing


DECISION MAKING

8/9/2019 ie256p3

48/54

[email protected]


E(pe$t Syste's /ES0

Replicates !ecision 'a"ing p$ocess

Kno-le!ge $ep$esentation !esc$ibes t+e -ay ane(pe$t -ol! app$oac+ t+e p$oble'

DECISION MAKING

8/9/2019 ie256p3

49/54

[email protected]

*ompetition nee!s very ast !ecisions an!

rapi! !evelopment o inormation systems.Concent$ate on -+at to !o $at+e$ t+an +o- to !o1

7or many companies6 inormation systems

cost 9: percent o overall costs.

DECISION MAKING

8/9/2019 ie256p3

50/54

[email protected]

CASE% Co'pte$ Ai!e! Soft-a$e Enginee$ing Tools

Soft-a$e Tools se! to ato'ate Soft-a$e Develop'ent &ife Cycle1

DECISION MAKING

4n!erstan!in /elational Batabases (usiness #e)uirements

8/9/2019 ie256p3

51/54

[email protected]

This type of Software

Development Life Cycle is

called Waterfall Model. Sinceit is difficult to swim up to the

waterfall stream it is costly to

go !ac" to the previous stages

in life cycle.

Therefore it is essential to

finish a good data model

!efore starting data!ase

design.

ST#$T%&'

(usiness #e)uirements

$*$L'S+S

D%S+&*

D,CUM%*T$T+,*(U+LD

T#$*S+T+,*

P#,DUCT+,*

,perational System

Software Development

Life Cycle!aterfall "o#el$

DECISION MAKING

8/9/2019 ie256p3

52/54

[email protected]

T+e #se of CASE in O$gani2ations

Objectives of CASE I'p$ove *ality of syste's !evelope!

Inc$ease spee! of !evelop'ent an! !esign

Ease an! i'p$ove testing p$ocess t+$og+ ato'ate! c+ec"ing

I'p$ove integ$ation of !evelop'ent activities via co''on'et+o!ologies

I'p$ove *ality an! co'pleteness of !oc'entation

5elp stan!a$!i2e t+e !evelop'ent p$ocess

I'p$ove p$oject 'anage'ent

Si'ply p$og$a' 'aintenance P$o'ote $esability

I'p$ove soft-a$e po$tability

DECISION MAKING

8/9/2019 ie256p3

53/54

[email protected]

#ect o 3anaement 5evel on Becisions

Manage'ent N'be$ of Cost of Ma"ing Info$'ation

&evel Decisions Poo$ Decisions Nee!s

Top &east 5ig+est St$ategic

Mi!!le Inte$'e!iate Inte$'e!iate I'ple'entation

4i$st>&ine Most &o-est Ope$ational

DECISION MAKING

8/9/2019 ie256p3

54/54

0mplementation

Decisions a$e seless nless t+ey a$e pt into

p$actice1

Co$age is t+e -illingness to sb'e$ge

oneself in t+e loneliness) t+e an(iety) an! t+e

gilt of a !ecision 'a"e$1

ie256p3

Documents