« Submodular Optimization and Approximation Algorithms

Satoru Iwata

Submodular functions are discrete analogues of convex functions. Examples include cut capacity functions, matroid rank functions, and entropy functions. Submodular functions can be minimized in polynomial time, which provides a fairly general framework of efficiently solvable combinatorial optimization problems. In contrast, the maximization problems are NP-hard and several approximation algorithms have been developed so far. In this lecture, I will review the above results in submodular optimization and present recent approximation algorithms for combinatorial optimization problems described in terms of submodular functions.

Scroll with j/k | | | Size

Slide: Submodular Optimization and Approximation Algorithm
Satoru Iwata   RIMS, Kyoto University

Submodular Optimization and Approximation Algorithm
Satoru Iwata RIMS, Kyoto University

1

Slide: Outline
 Submodular Functions Examples Discrete Convexity  Submodular Function Minimization  Approximation Algorithms  Submodular Function Maximization  Approximating Submodular Functions

Outline
Submodular Functions Examples Discrete Convexity Submodular Function Minimization Approximation Algorithms Submodular Function Maximization Approximating Submodular Functions

2

Slide: Submodular Functions
V:
Finite Set
V

f :2 R

X , Y  V

f ( X )  f (Y )  f ( X  Y )  f ( X  Y )
 Cut Capacity Functions  Matroid Rank Functions  Entropy Functions

X

Y

V

Submodular Functions
V:
Finite Set
V

f :2 R

X , Y V

f ( X ) f (Y ) f ( X Y ) f ( X Y )
Cut Capacity Functions Matroid Rank Functions Entropy Functions

X

Y

V

3

Slide: Cut Capacity Function
Cut Capacity

 ( X )  {c(a) | a : leaving X }
c( a )  0

s

t

X

Max Flow ValueMin Cut Capacity

Cut Capacity Function
Cut Capacity

( X ) {c(a) | a : leaving X }
c( a ) 0

s

t

X

Max Flow ValueMin Cut Capacity

4

Slide: Matroid Rank Functions
Matrix Rank Function Whitney (1935)

 ( X )  rank A[U , X ]
X

A
V

U

X  V ,  ( X ) | X | X  Y   ( X )   (Y )  : Submodular

Matroid Rank Functions
Matrix Rank Function Whitney (1935)

( X ) rank A[U , X ]
X

A
V

U

X V , ( X ) | X | X Y ( X ) (Y ) : Submodular

5

Slide: Entropy Functions
Information Sources

h( )  0

h(X ) : Entropy of the Joint Distribution
h( X )  h(Y )  h( X  Y )  h( X  Y )
X
Conditional Mutual Information

0

Entropy Functions
Information Sources

h( ) 0

h(X ) : Entropy of the Joint Distribution
h( X ) h(Y ) h( X Y ) h( X Y )
X
Conditional Mutual Information

0

6

Slide: Positive Definite Symmetric Matrices
X

f ( )  0
A
A[X ] f ( X )  log det A[ X ]
Ky Fans Inequality

f ( X )  f (Y )  f ( X  Y )  f ( X  Y )
Extension of the Hadamard Inequality
det A   Aii
iV

Positive Definite Symmetric Matrices
X

f ( ) 0
A
A[X ] f ( X ) log det A[ X ]
Ky Fans Inequality

f ( X ) f (Y ) f ( X Y ) f ( X Y )
Extension of the Hadamard Inequality
det A Aii
iV

7

Slide: Discrete Concavity
S T  f ( S  {u})  f ( S )  f (T  {u})  f (T )
Diminishing Returns

T

S

u

V

Discrete Concavity
S T f ( S {u}) f ( S ) f (T {u}) f (T )
Diminishing Returns

T

S

u

V

8

Slide: Discrete Convexity
Convex Function
f ( X )  f (Y )  f ( X  Y )  f ( X  Y )

y

X

Y

x

V

Discrete Convexity
Convex Function
f ( X ) f (Y ) f ( X Y ) f ( X Y )

y

X

Y

x

V

9

Slide: Discrete Convexity
V  {a, b, c}
{a, b}
{b, c}
Lovsz (1983)

f  Linear Interpolation

{b}

f  Convex
{c}

{a}



f  Submodular

 [0,1] : Chosen Uniformly at Random  f ( x)  E[ f ( X )] X : {v | x(v)   },

Discrete Convexity
V {a, b, c}
{a, b}
{b, c}
Lovsz (1983)

f Linear Interpolation

{b}

f Convex
{c}

{a}

f Submodular

[0,1] : Chosen Uniformly at Random f ( x) E[ f ( X )] X : {v | x(v) },

10

Slide: Submodular Function Minimization
Assumption: f ( )  0

Minimization Algorithm

X

Evaluation Oracle

f (X )

Minimizer

min{ f (Y ) | Y  V }?

Submodular Function Minimization
Assumption: f ( ) 0

Minimization Algorithm

X

Evaluation Oracle

f (X )

Minimizer

min{ f (Y ) | Y V }?

11

Slide: Submodular Function Minimization
Grtschel, Lovsz, Schrijver (1981, 1988)
Ellipsoid Method

O(n5 log M ) O(n7 log n)

Cunningham (1985)

O( n 7  n 8 )
Schrijver (2000)

Iwata, Fleischer, Fujishige (2000)

 : Time for Function Evaluation

M  max | f ( X ) |
X V

Submodular Function Minimization
Grtschel, Lovsz, Schrijver (1981, 1988)
Ellipsoid Method

O(n5 log M ) O(n7 log n)

Cunningham (1985)

O( n 7 n 8 )
Schrijver (2000)

Iwata, Fleischer, Fujishige (2000)

: Time for Function Evaluation

M max | f ( X ) |
X V

12

Slide: Base Polyhedra
R  {x | V  R}
V

x(Y )   x(v)
vY

Submodular Polyhedron

P( f )  {x | x  R , Y  V , x(Y )  f (Y )}
V

Base Polyhedron

B( f )  {x | x  P( f ), x(V )  f (V )}

Base Polyhedra
R {x | V R}
V

x(Y ) x(v)
vY

Submodular Polyhedron

P( f ) {x | x R , Y V , x(Y ) f (Y )}
V

Base Polyhedron

B( f ) {x | x P( f ), x(V ) f (V )}

13

Slide: Greedy Algorithm
L(v)

Edmonds (1970) Shapley (1971)

v1

v2

v

vn

y(v)  f ( L(v))  f ( L(v)  {v})

(v V )

y : Extreme Base
1 1    1 0 1  1     0  y (v1 )   f ( L(v1 ))    y (v )  f ( L(v ))   2   2     0         1  y (vn )  f ( L(vn ))

Greedy Algorithm
L(v)

Edmonds (1970) Shapley (1971)

v1

v2

v

vn

y(v) f ( L(v)) f ( L(v) {v})

(v V )

y : Extreme Base
1 1 1 0 1 1 0 y (v1 ) f ( L(v1 )) y (v ) f ( L(v )) 2 2 0 1 y (vn ) f ( L(vn ))

14

Slide: Min-Max Theorem
Theorem
 Y V

Edmonds (1970)

min f (Y )  max{ x (V ) | x  B( f )}

x (v) : min{0, x(v)}
x (V )  x(Y )  f (Y )

Min-Max Theorem
Theorem
Y V

Edmonds (1970)

min f (Y ) max{ x (V ) | x B( f )}

x (v) : min{0, x(v)}
x (V ) x(Y ) f (Y )

15

Slide: Combinatorial Approach
Extreme Base yL  B( f )
Convex Combination

x   L y L
L

x

Cunningham (1985)

O(n6 M log nM )

M  max | f ( X ) |
X V

Combinatorial Approach
Extreme Base yL B( f )
Convex Combination

x L y L
L

x

Cunningham (1985)

O(n6 M log nM )

M max | f ( X ) |
X V

16

Slide: Combinatorial Approach
x   L y L
L

L

y L : Extreme Base

x(v)  0, v  T x(v)  0, v  T

T

yL (T )  f (T ), L  .  x(T )  f (T ).
x (V )  x(T )  f (T )


T : Minimizer

Combinatorial Approach
x L y L
L

L

y L : Extreme Base

x(v) 0, v T x(v) 0, v T

T

yL (T ) f (T ), L . x(T ) f (T ).
x (V ) x(T ) f (T )

T : Minimizer

17

Slide: Submodular Function Minimization
Grtschel, Lovsz, Schrijver (1981, 1988)
Ellipsoid Method

O(n5 log M ) O(n7 log n)

Cunningham (1985)

O( n 7  n 8 )
Schrijver (2000)
Fleischer, Iwata (2000)

Iwata, Fleischer, Fujishige (2000)

Iwata (2002)
Fully Combinatorial

Iwata (2003)

Orlin (2007)

O((n 4  n5 ) log M )
Iwata, Orlin (2009)

O(n5  n6 )

Submodular Function Minimization
Grtschel, Lovsz, Schrijver (1981, 1988)
Ellipsoid Method

O(n5 log M ) O(n7 log n)

Cunningham (1985)

O( n 7 n 8 )
Schrijver (2000)
Fleischer, Iwata (2000)

Iwata, Fleischer, Fujishige (2000)

Iwata (2002)
Fully Combinatorial

Iwata (2003)

Orlin (2007)

O((n 4 n5 ) log M )
Iwata, Orlin (2009)

O(n5 n6 )

18

Slide: The Fujishige-Wolfe Algorithm
Minimize x
2

Fujishige (1984)

subject to x  B( f )

x : opt.sol.
S : {v | x (v)  0}
T : {v | x (v)  0}




x




Minimal Minimizer Maximal Minimizer

f (S )  f (T )  min f ( X )
X V

The Fujishige-Wolfe Algorithm
Minimize x
2

Fujishige (1984)

subject to x B( f )

x : opt.sol.
S : {v | x (v) 0}
T : {v | x (v) 0}

x

Minimal Minimizer Maximal Minimizer

f (S ) f (T ) min f ( X )
X V

19

Slide: Evacuation Problem Dynamic Flow
S
T
Hoppe, Tardos (2000)

c(a) : Capacity  (a) : Transit Time
b(v) : Supply/Demand

o(X ) : Maximum Amount of Flow from X  S to T \ X.
Feasible

b( X )  o( X ), X  S  T

Evacuation Problem Dynamic Flow
S
T
Hoppe, Tardos (2000)

c(a) : Capacity (a) : Transit Time
b(v) : Supply/Demand

o(X ) : Maximum Amount of Flow from X S to T \ X.
Feasible

b( X ) o( X ), X S T

20

Slide: Multiterminal Source Coding
RY

X

Encoder
Encoder

RX

Y

RY

Decoder

H ( X ,Y )
H (Y )

H (Y | X )

Slepian, Wolf (1973)
H(X | Y)
H (X ) H ( X , Y )

RX

Multiterminal Source Coding
RY

X

Encoder
Encoder

RX

Y

RY

Decoder

H ( X ,Y )
H (Y )

H (Y | X )

Slepian, Wolf (1973)
H(X | Y)
H (X ) H ( X , Y )

RX

21

Slide: Multiclass Queueing Systems
Arrival Interval
Server

Service Time

Arrival Rate i Multiclass M/M/1 Preemptive

 i Service Rate
Control Policy

Multiclass Queueing Systems
Arrival Interval
Server

Service Time

Arrival Rate i Multiclass M/M/1 Preemptive

i Service Rate
Control Policy

22

Slide: Performance Region
s j : Expected Staying Time of a Job in j

s : Achievable
  i Si
iX

i : i i ,
sj
RY
s:


iV

i

1

   1  
iX i iX

i

, X  V

i

Coffman, Mitrani (1980)

si

Performance Region
s j : Expected Staying Time of a Job in j

s : Achievable
i Si
iX

i : i i ,
sj
RY
s:

iV

i

1

1
iX i iX

i

, X V

i

Coffman, Mitrani (1980)

si

23

Slide: A Class of Submodular Functions
x, y, z  RV 
Itoko & Iwata (2007)

h : Nonnegative, Nondecreasing, Convex

f ( X )  z( X )  y( X )h( x( X ))
Submodular

(X V )

  i Si
iX

   1  
iX i iX

i

zi : i Si
, X  V

i yi : i
1 h( x) : 1 x

i

xi : i

A Class of Submodular Functions
x, y, z RV
Itoko & Iwata (2007)

h : Nonnegative, Nondecreasing, Convex

f ( X ) z( X ) y( X )h( x( X ))
Submodular

(X V )

i Si
iX

1
iX i iX

i

zi : i Si
, X V

i yi : i
1 h( x) : 1 x

i

xi : i

24

Slide: Zonotope in 3D
w( X )  ( x( X ), y( X ), z( X ))

z

Z  conv{w( X ) | X  V }
Zonotope

~ f ( x, y, z )  z  yh( x)
min{ f ( X ) | X  V } ~  min{ f ( x, y, z ) | ( x, y, z ) : Lower Extreme Point of Z }

~ Remark: f ( x, y, z ) is NOT concave!

Zonotope in 3D
w( X ) ( x( X ), y( X ), z( X ))

z

Z conv{w( X ) | X V }
Zonotope

~ f ( x, y, z ) z yh( x)
min{ f ( X ) | X V } ~ min{ f ( x, y, z ) | ( x, y, z ) : Lower Extreme Point of Z }

~ Remark: f ( x, y, z ) is NOT concave!

25

Slide: Line Arrangement

xi  yi   zi



Enumerating All the Cells Topological Sweeping Method Edelsbrunner, Guibas (1989)

O( n 2 )

Line Arrangement

xi yi zi

Enumerating All the Cells Topological Sweeping Method Edelsbrunner, Guibas (1989)

O( n 2 )

26

Slide: Symmetric Submodular Functions
f :2 R
V

Symmetric

f ( X )  f (V \ X ),

X  V .

Symmetric Submodular Function Minimization

min{ f ( X ) |   X  V }?
O(n3 ) Queyranne (1998)
Minimum Cut Algorithm by MA-ordering Nagamochi & Ibaraki (1992) Minimum Degree Ordering Nagamochi (2007)

Symmetric Submodular Functions
f :2 R
V

Symmetric

f ( X ) f (V \ X ),

X V .

Symmetric Submodular Function Minimization

min{ f ( X ) | X V }?
O(n3 ) Queyranne (1998)
Minimum Cut Algorithm by MA-ordering Nagamochi & Ibaraki (1992) Minimum Degree Ordering Nagamochi (2007)

27

Slide: Submodular Partition
Minimize subject to

 f (V )
i 1 i

k

Greedy Split

V  V1  Vk
Vi V j   (i  j )

f : Monotone or Symmetric

2(1  1 k )-Approximation
Zhao, Nagamochi, Ibaraki (2005)

Submodular Partition
Minimize subject to

f (V )
i 1 i

k

Greedy Split

V V1 Vk
Vi V j (i j )

f : Monotone or Symmetric

2(1 1 k )-Approximation
Zhao, Nagamochi, Ibaraki (2005)

28

Slide: Questions
What Kind of Approximation Algorithms Can Be Extended to Optimization Problems with Submodular Cost or Constraints ?
Cf. Submodular Flow (Edmonds & Giles, 1977)

How Can We Exploit Discrete Convexity in Design of Approximation Algorithms?
Cf. Ellipsoid Method (Grtschel, Lovsz & Schrijver, 1981)

Questions
What Kind of Approximation Algorithms Can Be Extended to Optimization Problems with Submodular Cost or Constraints ?
Cf. Submodular Flow (Edmonds & Giles, 1977)

How Can We Exploit Discrete Convexity in Design of Approximation Algorithms?
Cf. Ellipsoid Method (Grtschel, Lovsz & Schrijver, 1981)

29

Slide: Submodular Vertex Cover
Graph G  (V , E )
Submodular Function
f : 2V  R 

Find a Vertex Cover S  V Minimizing f (S ) 2-Approximation Algorithm
Goel, Karande, Tripathi, Wang (FOCS 2009) Iwata & Nagano (FOCS 2009)

Submodular Vertex Cover
Graph G (V , E )
Submodular Function
f : 2V R

Find a Vertex Cover S V Minimizing f (S ) 2-Approximation Algorithm
Goel, Karande, Tripathi, Wang (FOCS 2009) Iwata & Nagano (FOCS 2009)

30

Slide: Relaxation Problem
Convex Programming Relaxation (CPR)

Minimize

 f ( x)

subject to x(u)  x(v)  1 (e  (u, v)  E)
x(v)  0 (v V )

CPR has a half-integral optimal solution. 2-Approximation Algorithm

Relaxation Problem
Convex Programming Relaxation (CPR)

Minimize

f ( x)

subject to x(u) x(v) 1 (e (u, v) E)
x(v) 0 (v V )

CPR has a half-integral optimal solution. 2-Approximation Algorithm

31

Slide: Submodular Cost Set Cover
Find X  N Covering U with Minimum f (X ).
N

U

 : max | Nu |
uU

 -Approximation Rounding Algorithm Primal-Dual Algorithm

Nu

u

Submodular Cost Set Cover
Find X N Covering U with Minimum f (X ).
N

U

: max | Nu |
uU

-Approximation Rounding Algorithm Primal-Dual Algorithm

Nu

u

32

Slide: Submodular Multiway Partition
Chekuri & Ene (FOCS 2011)

Minimize

 f (V )
i 1 i

k

subject to V  V1   Vk

Vi V j   (i  j )

si Vi

(i  1,..., k )

f : Nonnegative Submodular

Submodular Multiway Partition
Chekuri & Ene (FOCS 2011)

Minimize

f (V )
i 1 i

k

subject to V V1 Vk

Vi V j (i j )

si Vi

(i 1,..., k )

f : Nonnegative Submodular

33

Slide: Relaxation Problem
Convex Programming Relaxation

Minimize
subject to


i 1 k

k

 f ( xi )
i

 x (v )  1
i 1

(v V )

xi (si )  1 (i  1,..., k )
xi (v)  0
Ellipsoid Method


(i  1,..., k , v V )

xi : Optimal Solution

Relaxation Problem
Convex Programming Relaxation

Minimize
subject to

i 1 k

k

f ( xi )
i

x (v ) 1
i 1

(v V )

xi (si ) 1 (i 1,..., k )
xi (v) 0
Ellipsoid Method

(i 1,..., k , v V )

xi : Optimal Solution

34

Slide: Rounding Scheme
f : Symmetric Submodular
 [0,1] : Chosen Uniformly at Random
Ai : {v | xi (v)   } (i  1,..., k ),
Uncross ( A1 ,..., Ak )

U : V   Ai
i 1

k

Return ( A1 ,..., Ak 1 , Ak U ) 3/2-Approximate Solution

Rounding Scheme
f : Symmetric Submodular
[0,1] : Chosen Uniformly at Random
Ai : {v | xi (v) } (i 1,..., k ),
Uncross ( A1 ,..., Ak )

U : V Ai
i 1

k

Return ( A1 ,..., Ak 1 , Ak U ) 3/2-Approximate Solution

35

Slide: Rounding Scheme
f : Nonnegative Submodular
  ( 1 ,1] : Chosen Uniformly at Random 2
Ai : {v | xi (v)   } (i  1,..., k ),
Return ( A1 ,..., Ak 1 , Ak U )
2-Approximate Solution Improvement over the (k  1) -Approximation by Zhao, Nagamochi, & Ibaraki (2005)

U : V   Ai
i 1

k

Rounding Scheme
f : Nonnegative Submodular
( 1 ,1] : Chosen Uniformly at Random 2
Ai : {v | xi (v) } (i 1,..., k ),
Return ( A1 ,..., Ak 1 , Ak U )
2-Approximate Solution Improvement over the (k 1) -Approximation by Zhao, Nagamochi, & Ibaraki (2005)

U : V Ai
i 1

k

36

Slide: Submodular Function Maximization
Monotone Submodular Functions
Nemhauser, Wolsey, Fisher (1978) Cardinality Constraint (1-1/e)-Approximation

Calinescu, Chekuri, Pl, Vondrk (IPCO 2007) Vondrk (STOC 2008) Filmes and Ward (FOCS 2012) Matroid Constraint (1-1/e)-Approximation

Submodular Function Maximization
Monotone Submodular Functions
Nemhauser, Wolsey, Fisher (1978) Cardinality Constraint (1-1/e)-Approximation

Calinescu, Chekuri, Pl, Vondrk (IPCO 2007) Vondrk (STOC 2008) Filmes and Ward (FOCS 2012) Matroid Constraint (1-1/e)-Approximation

37

Slide: Greedy Approximation Algorithm
Nemhauser, Wolsey, Fisher (1978)

f : Monotone Submodular Function
Maximize f (S ) subject to | S | k Greedy Algorithm
v1
v j 1

T0 :  T j : T j 1  {v j }

T j 1

v j : arg max f (T j 1  {v})

j  1,..., k

Greedy Approximation Algorithm
Nemhauser, Wolsey, Fisher (1978)

f : Monotone Submodular Function
Maximize f (S ) subject to | S | k Greedy Algorithm
v1
v j 1

T0 : T j : T j 1 {v j }

T j 1

v j : arg max f (T j 1 {v})

j 1,..., k

38

Slide: Greedy Approximation Algorithm
S : Optimal Solution  : f (S )
*
*

  k j   i ( j  1,..., k )
i 1

j 1

 j : f (T j )  f (T j 1 )

f (Tk )   i
i 1

k

) f ( S )  f ( S  T j 1 )
* *

 f (T j 1 ) 

uS * \T j 1

  f (T

j 1

 {u})  f (T j 1 )
j 1



 f (T j 1 )  k j


i 1

i

Greedy Approximation Algorithm
S : Optimal Solution : f (S )
*
*

k j i ( j 1,..., k )
i 1

j 1

j : f (T j ) f (T j 1 )

f (Tk ) i
i 1

k

) f ( S ) f ( S T j 1 )
* *

f (T j 1 )

uS * \T j 1

f (T

j 1

{u}) f (T j 1 )
j 1

f (T j 1 ) k j

i 1

i

39

Slide: Greedy Approximation Algorithm
Minimize


i 1

k

i


i 1

k

i

 f (Tk )
*

subject to

k 0  0   1    1 k           2         0            1  1 k    k   

 : f (S )
  j :



1 1   k k

j

j  0
k

( j  1,..., k )

( j  1,..., k )

  1 k   1   i   1  1  k     1  e      i 1

Greedy Approximation Algorithm
Minimize

i 1

k

i

i 1

k

i

f (Tk )
*

subject to

k 0 0 1 1 k 2 0 1 1 k k

: f (S )
j :

1 1 k k

j

j 0
k

( j 1,..., k )

( j 1,..., k )

1 k 1 i 1 1 k 1 e i 1

40

Slide: Submodular Welfare Problem
Utility Functions

f1 ,..., f k (Monotone, Submodular)

Maximize subject to

 f (V )
i 1 i i

k

V  V1  Vk
Vi V j   (i  j )

(1  1 e) -Approximation

Vondrk (2008)

Submodular Welfare Problem
Utility Functions

f1 ,..., f k (Monotone, Submodular)

Maximize subject to

f (V )
i 1 i i

k

V V1 Vk
Vi V j (i j )

(1 1 e) -Approximation

Vondrk (2008)

41

Slide: Submodular Function Maximization
Nonnegative Submodular Functions
Feige, Mirrokni, Vondrk (FOCS 2007) 2/5-Approximation Lee, Mirrokni, Nagarajan, Sviridenko (STOC 2009) 1/4-Approximation (Matroid Constraint) 1/5-Approximation (Knapsack Constraints) Buchbinder, Feldman, Naor, Schwartz (FOCS 2012) 1/2-Approximation

Submodular Function Maximization
Nonnegative Submodular Functions
Feige, Mirrokni, Vondrk (FOCS 2007) 2/5-Approximation Lee, Mirrokni, Nagarajan, Sviridenko (STOC 2009) 1/4-Approximation (Matroid Constraint) 1/5-Approximation (Knapsack Constraints) Buchbinder, Feldman, Naor, Schwartz (FOCS 2012) 1/2-Approximation

42

Slide: Double-Greedy Algorithm
Buchbinder, Feldman, Naor, Schwartz (FOCS 2012)

Initialize:

A :  , B : V , D : . A  B, D  B \ A.

Iteration: grow A or shrink B. Invariants: Output: either A or B that has the greater value.

A B
V

Double-Greedy Algorithm
Buchbinder, Feldman, Naor, Schwartz (FOCS 2012)

Initialize:

A : , B : V , D : . A B, D B \ A.

Iteration: grow A or shrink B. Invariants: Output: either A or B that has the greater value.

A B
V

43

Slide: Double-Greedy Algorithm
While

A D  B Select u  B \ ( D  A).  : max{ f ( A {u})  f ( A),0}.  : max{ f ( B \ {u})  f ( B),0}.
If

    0,

then
 
  

A : A {u} with probability B : B \ {u} with probability
Else



,

.

D : D {u}.

Double-Greedy Algorithm
While

A D B Select u B \ ( D A). : max{ f ( A {u}) f ( A),0}. : max{ f ( B \ {u}) f ( B),0}.
If

0,

then

A : A {u} with probability B : B \ {u} with probability
Else

,

.

D : D {u}.

44

Slide: Double-Greedy Algorithm
The expectation

E[ f ( A)  f ( B)  2 f (Z )]

never decreases in the process.

) Expected changes of f ( A)  f ( B)  2 f (Z )
at each iteration is at least

2 (   )       0.        
2

Double-Greedy Algorithm
The expectation

E[ f ( A) f ( B) 2 f (Z )]

never decreases in the process.

) Expected changes of f ( A) f ( B) 2 f (Z )
at each iteration is at least

2 ( ) 0.
2

45

Slide: Approximating Submodular Functions
X
Algorithm

f (X )

Evaluation Oracle

Function

f

 f (Y )

Y

Approximating Submodular Functions
X
Algorithm

f (X )

Evaluation Oracle

Function

f

f (Y )

Y

46

Slide: Approximating Submodular Functions
Goemans, Harvey, Iwata & Mirrokni (SODA 2009)

f ( )  0, f ( X )  0, X  V .  Construct a set function f such that
  f ( X )  f ( X )   (n) f ( X ), X  V .
For what function  is this possible? Algorithm with  (n)  O( n log n) for monotone submodular functions

Approximating Submodular Functions
Goemans, Harvey, Iwata & Mirrokni (SODA 2009)

f ( ) 0, f ( X ) 0, X V . Construct a set function f such that
f ( X ) f ( X ) (n) f ( X ), X V .
For what function is this possible? Algorithm with (n) O( n log n) for monotone submodular functions

47

Slide: Ellipsoidal Approximation
K : Centrally Symmetric Convex Body (x  K  x  K )

E (A) : Maximum Volume Inscribed Ellipsoid
(The John Ellipsoid)

E ( A)  K  n E ( A)

Ellipsoidal Approximation
K : Centrally Symmetric Convex Body (x K x K )

E (A) : Maximum Volume Inscribed Ellipsoid
(The John Ellipsoid)

E ( A) K n E ( A)

48

Slide: Submodular Load Balancing
Svitkina & Fleischer (FOCS 2008)

f1 ,..., f m : Monotone Submodular Functions
{V1 ,..., m } V

min max f j (V j ) ?
j

f j ( X ) :  p j (v)
vX

Scheduling 2-Approximation Algorithm Lenstra, Shmoys, Tardos (1990)

O( n log n) -Approximation Algorithm

Submodular Load Balancing
Svitkina & Fleischer (FOCS 2008)

f1 ,..., f m : Monotone Submodular Functions
{V1 ,..., m } V

min max f j (V j ) ?
j

f j ( X ) : p j (v)
vX

Scheduling 2-Approximation Algorithm Lenstra, Shmoys, Tardos (1990)

O( n log n) -Approximation Algorithm

49

Slide: Learning Submodular Functions
Balcan & Harvey (STOC 2010)

PMAC-learning

 Construct a set function f such that   f ( X )  f ( X )   (n) f ( X )
holds for most (1   fraction) of X . For what function  is this possible with probability 1   in poly(n,1 /  ,1 /  ) time?

Learning Submodular Functions
Balcan & Harvey (STOC 2010)

PMAC-learning

Construct a set function f such that f ( X ) f ( X ) (n) f ( X )
holds for most (1 fraction) of X . For what function is this possible with probability 1 in poly(n,1 / ,1 / ) time?

50

Slide: Summary
       Submodular Functions Arise Everywhere. Discrete Analogue of Convexity. Combinatorial Algorithms for Minimization. Exploit Special Structures of Problems. Approx. Algorithms by Discrete Convexity Approx. Algorithms for Maximization Learning Submodular Functions

Summary
Submodular Functions Arise Everywhere. Discrete Analogue of Convexity. Combinatorial Algorithms for Minimization. Exploit Special Structures of Problems. Approx. Algorithms by Discrete Convexity Approx. Algorithms for Maximization Learning Submodular Functions

51