lecture 17 - nlp - formulations

Upload: willa-catherine

Post on 06-Apr-2018

216 views

Category:

Documents


0 download

TRANSCRIPT

  • 8/3/2019 Lecture 17 - NLP - Formulations

    1/18

    BUS321

    Management Science

    Lecture

    NLP - Modeling Non-linear Progra

    Copyright @ Jiong Sun

    BUS321001BUS321001::ManagementScienceManagementScience

    Lecture 17:Non-Linear Programming

    Modeling Examples

    2

    Objectives

    Objectives

    Know how to model non-linear programs (NLPs)(Ch. 11.2)

    Applications of NLP to location problems, portfoliomanagement, and regression

  • 8/3/2019 Lecture 17 - NLP - Formulations

    2/18

    BUS321

    Management Science

    Lecture

    NLP - Modeling Non-linear Progra

    Copyright @ Jiong Sun

    Linear Programming Model

    1 1 2 2

    11 1 12 2 1n n 1

    21 1 22 2 2n n 2

    m1 1 2 2 mn n m

    Maximize .....

    subject to

    a x + a x + ... +a x b

    a x + a x + ... +a x b

    a x + a x + ... +a x b

    n n

    m

    c x c x c x

    x

    + + +

    1 2, , ..., 0

    nx x

    ASSUMPTIONS:

    Proportionality Assumption

    Objective function

    Constraints

    Additivity Assumption

    Objective function

    Constraints

    Wha

    tis

    a

    Non-Linear

    Prog

    ram?

    maximize 3 sin x + xy + y3 - 3z + log zSubject to x2 + y2 = 1

    x + 4z 2z 0

    A non-linear program is permitted to have non-linear constraints or objectives.

    A linear program is a special case of non-linearprogramming!

  • 8/3/2019 Lecture 17 - NLP - Formulations

    3/18

    BUS321

    Management Science

    Lecture

    NLP - Modeling Non-linear Progra

    Copyright @ Jiong Sun

    Non-Linear Programs(NLP)

    Nonlinear objective function f(x) and/orNonlinear constraints gi(x).

    Today: we will present several types of non-linearprograms.

    ( )1 2, , ,

    ( )

    ( ) , 1, 2, ,

    n

    i i

    Let x x x x

    Max f x

    g x b i m

    =

    =

    Unconstrained Facility Location

    0

    2

    4

    6

    8

    10

    12

    14

    16

    y

    0 2 4 6 8 10 12 14 16

    C (2)

    (7)

    B

    A (19)

    P ?

    D (5)

    x

    Loc. Demand

    A: (8,2) 19

    B: (3,10) 7

    C: (8,15) 2

    D: (14,13) 5

    P: ?

    This is the warehouse location problem with a single

    warehouse that can be located anywhere in the plane.

  • 8/3/2019 Lecture 17 - NLP - Formulations

    4/18

    BUS321

    Management Science

    Lecture

    NLP - Modeling Non-linear Progra

    Copyright @ Jiong Sun

    Costs proportional to distance;known daily demands.

    An NLP

    2 28 2( ) ( )x y + d(P,A) =

    2 214 13( ) ( )x y + d(P,D) =

    minimize 19 d(P,A) + + 5 d(P,D)

    subject to: P is unconstrained

    Herearethe objective valuesfor55

    differentlocations.

    0

    50

    100

    150

    200

    250

    300

    350

    values

    for y

    Objectivevalue

    x = 0

    x = 2

    x = 4

    x = 6

    x = 8

    x = 10

    x = 12

  • 8/3/2019 Lecture 17 - NLP - Formulations

    5/18

    BUS321

    Management Science

    Lecture

    NLP - Modeling Non-linear Progra

    Copyright @ Jiong Sun

    Facil ity Location. What happensif P must be

    within aspecified region?

    02

    4

    6

    8

    10

    12

    14

    16

    y

    0 2 4 6 8 10 12 14 16

    C (2)

    (7)

    B

    A (19)

    P ?

    D (5)

    x

    Themodel

    2 219 8 2( ) ( )x y +

    2 25 14 13( ) ( )x y +

    + +Minimize

    Subject to x 75 y 11

    x + y 24

  • 8/3/2019 Lecture 17 - NLP - Formulations

    6/18

    BUS321

    Management Science

    Lecture

    NLP - Modeling Non-linear Progra

    Copyright @ Jiong Sun

    0-1integer programsas NLPs

    minimize Sj cj xj

    subject to Sj aij xj = bi for all ixj is 0 or 1 for all j

    is nearly equivalent to

    minimize Sj cj xj + 106 Sj xj (1- xj).

    subject to Sj aij xj = bi for all i

    0 xj 1 for all j

    Someco

    mment

    s

    on non-linear

    m

    odels

    The fact that non-linear models can model somuch is perhaps a bad sign How can we solve non-linear programs if we have

    trouble with integer programs?

    Recall, in solving integer programs we usetechniques that rely on the integrality.

    Fact: some non-linear models can be solved,and some are WAY too difficult to solve. More

    on this later.

  • 8/3/2019 Lecture 17 - NLP - Formulations

    7/18

    BUS321

    Management Science

    Lecture

    NLP - Modeling Non-linear Progra

    Copyright @ Jiong Sun

    Another Example: Machine Values

    Buy a machine and keep it for t years, andthen sell it. (0 t 10)

    All values are measured in $ million

    Cost of machine = 1.5

    Revenue = 4(1 - .75t)

    Salvage value = 1/(1 + t)

    Machine values

    00.5

    1

    1.5

    2

    2.5

    3

    3.5

    4

    4.5

    0.

    2 1

    1.

    8

    2.

    6

    3.

    4

    4.

    2 5

    5.

    8

    6.

    6

    7.

    4

    8.

    2 9

    9.

    8

    Time

    Millionsofdollars

    revenue

    salvage

    total

    How longs

    hou

    ld wekeep thema

    chine?

  • 8/3/2019 Lecture 17 - NLP - Formulations

    8/18

    BUS321

    Management Science

    Lecture

    NLP - Modeling Non-linear Progra

    Copyright @ Jiong Sun

    Non-linearities Because of T ime

    Discount rates Decreasing value of equipment over time

    wear and tear, improvements in technology

    Tax implications

    Salvage value

    Non-linear

    ities

    in Pricing

    The price of an item may depend on the numbersold

    quantity discounts (small sellers)

    price elasticity (monopoly)

    Complex interactions because of substitutions: Lowering the price of GM automobiles will decrease the

    demand for the competitors

  • 8/3/2019 Lecture 17 - NLP - Formulations

    9/18

    BUS321

    Management Science

    Lecture

    NLP - Modeling Non-linear Progra

    Copyright @ Jiong Sun

    Non-linearities because ofcongestion

    The time it takes to go from Lincoln Park to IITby car depends non-linearly on the congestion.

    As congestion increases just to its limit, thetraffic sometimes comes to a near halt.

    Non-linearities because ofpenalties

    Consider any linear equality constraint:

    e.g., 3x1 + 5x2 + 4x3 = 17

    Suppose it is a soft constraint and we permit solutionsviolating it. We can then write:

    3x1 + 5x2 + 4x3 - y = 17

    And we may include a term of 10y2 in the objectivefunction.

    This adds flexibility to the solution by discouragingviolation of our goals

  • 8/3/2019 Lecture 17 - NLP - Formulations

    10/18

    BUS321

    Management Science

    Lecture

    NLP - Modeling Non-linear Progra

    Copyright @ Jiong Sun

    Portfolio Optimization

    In the following slides, we will show how tomodel portfolio optimization as NLPs

    The key concept is that risk can be modeledusing non-linear equations

    Since this is one of the most famous applicationsof non-linear programming, we cover it in muchmore detail

    Ris

    k vs.

    Retur

    n In finance, one needs to trade-off risk and

    return. For a given rate of return, one wants to minimize risk.

    For a given rate of risk, one wants to maximize return.

    Return is modeled as expected value.Risk is modeled as variance (or standarddeviation).

  • 8/3/2019 Lecture 17 - NLP - Formulations

    11/18

    BUS321

    Management Science

    Lecture

    NLP - Modeling Non-linear Progra

    Copyright @ Jiong Sun

    Expectations Add

    Suppose that X and Y are random variables

    E(X + Y) = E(X) + E(Y)

    Interpretation: Suppose that the expected return in one year for

    Stock 1 is 9%.

    Suppose that the expected return in one year forStock 2 is 10%

    If you put $100 in Stock 1, and $200 in Stock 2,your expected return is $9 + $20 = $29.

    Variances do notadd (atleast notsimply)

    Suppose that X and Y are random variables

    Var(aX + bY) =a2 Var(X) + b2 Var(Y) + 2ab Cov(X, Y)

    Example. The risk of investing in umbrellasand sunglasses is less than the risk of eitherinvestment by itself.

    In general:Var(X1 + X2 + + Xn) = 1 ( ) 2 ( , )

    n

    i i ji i jVar X Cov X X

    =