CS 332: Algorithms

Greedy Algorithms Continued

David Luebke 1
12/09/21
 Administrivia
 Hand in Homework 4
 Hand back remaining midterms

David Luebke 2
12/09/21
 Review:
LCS Via Dynamic Programming
 Longest common subsequence (LCS) problem:
 Given two sequences x[1..m] and y[1..n], find the
longest subsequence which occurs in both
 Brute-force algorithm: 2m subsequences of x to
check against n elements of y: O(n 2m)
 Define c[i,j] = length of LCS of x[1..i], y[1..j]
 Theorem:
c[i  1, j  1]  1 if x[i ]  y[ j ],
c[i, j ]  
 max(c[i, j  1], c[i  1, j ]) otherwise
David Luebke 3
12/09/21
Review: Structure Of LCS Problem
c[i  1, j  1]  1 if x[i ]  y[ j ],
c[i, j ]  
 max(c[i, j  1], c[i  1, j ]) otherwise
 Observation 1: Optimal substructure
 Thus a simple recursive algorithm will suffice
 Subproblems: LCS of prefixes of x,y
 Observation 2: Overlapping subproblems
 Many recurring instances of each subproblem
 Unlike divide and conquer, where subproblems are distinct

David Luebke 4
12/09/21
 Review: Memoization
 Memoization is one way to deal with overlapping
subproblems
 After computing the solution to a subproblem, store in
a table
 Subsequent calls just do a table lookup
 Can modify recursive alg to use memoziation:
 There are mn subproblems
 Each wanted 3 times: compute once, lookup twice
 Running time: O(mn). Running space: O(mn).

David Luebke 5
12/09/21
 Review: Dynamic Programming
 Dynamic programming: build table bottom-up
 Same table as memoization, but instead of starting
at (m,n) and recursing down, start at (1,1)
 Each cell can be computed in constant time from
its 3 neighbors to the left and up
 Running time/space: O(mn)
 Can actually reduce space to O(min(m,n)) by
keeping only two rows/columns around

David Luebke 6
12/09/21
 Review: Dynamic Programming
 Summary of the basic idea:
 Optimal substructure: optimal solution to problem
consists of optimal solutions to subproblems
 Overlapping subproblems: few subproblems in total,
many recurring instances of each
 Solve bottom-up, building a table of solved
subproblems that are used to solve larger ones
 Variations possible!
 “Table” could be 3-dimensional, triangular, a tree, etc.

David Luebke 7
12/09/21
 Greedy Algorithms
 A greedy algorithm always makes the choice that
looks best at the moment
 The hope: a locally optimal choice will lead to a globally
optimal solution
 For some problems, it works
 My everyday examples:
 Walking to the Corner
 Playing a bridge hand
 Dynamic programming can be overkill; greedy
algorithms tend to be easier to code

David Luebke 8
12/09/21
 Activity-Selection Problem
 Problem: get your money’s worth out of a
carnival
 Buy a wristband that lets you onto any ride
 Lots of rides, each starting and ending at different
times
 Your goal: ride as many rides as possible
 Another, alternative goal that we don’t solve here:
maximize time spent on rides
 Welcome to the activity selection problem

David Luebke 9
12/09/21
 Activity-Selection
 Formally:
 Given a set S of n activities
si = start time of activity i
fi = finish time of activity i
 Find max-size subset A of compatible activities
3
4 6
2
1 5

 Assume (wlog) that f1  f2  …  fn

David Luebke 10
12/09/21
 Activity Selection:
Optimal Substructure
 Let k be the minimum activity in A (i.e., the one
with the earliest finish time). Then A - {k} is an
optimal solution to S’ = {i  S: si  fk}
 In words: once activity #1 is selected, the problem
reduces to finding an optimal solution for activity-
selection over activities in S compatible with #1
 Proof: if we could find optimal solution B’ to S’ with |B|
> |A - {k}|,
 Then B  {k} is compatible
 And |B  {k}| > |A|

David Luebke 11
12/09/21
 Activity Selection:
Repeated Subproblems
 Consider a recursive algorithm that tries all
possible compatible subsets to find a maximal
set, and notice repeated subproblems:
S
1A?
yes no

S’ S-{1}
2A? 2A?
yes no yes no

S’’ S’-{2} S’’ S-{1,2}

David Luebke 12
12/09/21
 Greedy Choice Property
 Dynamic programming? Memoize? Yes, but…
 Activity selection problem also exhibits the greedy
choice property:
 Locally optimal choice  globally optimal sol’n
 Them 17.1: if S is an activity selection problem sorted
by finish time, then  optimal solution
A  S such that {1}  A
 Sketch of proof: if  optimal solution B that does not contain
{1}, can always replace first activity in B with {1} (Why?).
Same number of activities, thus optimal.

David Luebke 13
12/09/21
 Activity Selection:
A Greedy Algorithm
 So actual algorithm is simple:
 Sort the activities by finish time
 Schedule the first activity
 Then schedule the next activity in sorted list which
starts after previous activity finishes
 Repeat until no more activities
 Intuition is even more simple:
 Always pick the shortest ride available at the time

David Luebke 14
12/09/21
Minimum Spanning Tree Revisited
 Recall: MST problem has optimal substructure
 Prove it
 Is Prim’s algorithm greedy? Why?
 Is Kruskal’s algorithm greedy? Why?

David Luebke 15
12/09/21
Greedy Vs. Dynamic Programming:
The Knapsack Problem
 The famous knapsack problem:
 A thief breaks into a museum. Fabulous paintings,
sculptures, and jewels are everywhere. The thief has
a good eye for the value of these objects, and knows
that each will fetch hundreds or thousands of dollars
on the clandestine art collector’s market. But, the
thief has only brought a single knapsack to the scene
of the robbery, and can take away only what he can
carry. What items should the thief take to maximize
the haul?

David Luebke 16
12/09/21
 The Knapsack Problem
 More formally, the 0-1 knapsack problem:
 The thief must choose among n items, where the ith item
worth vi dollars and weighs wi pounds
 Carrying at most W pounds, maximize value
 Note: assume vi, wi, and W are all integers
 “0-1” b/c each item must be taken or left in entirety
 A variation, the fractional knapsack problem:
 Thief can take fractions of items
 Think of items in 0-1 problem as gold ingots, in fractional
problem as buckets of gold dust

David Luebke 17
12/09/21
 The Knapsack Problem:
Optimal Substructure
 Both variations exhibit optimal substructure
 To show this for the 0-1 problem, consider the
most valuable load weighing at most W pounds
 If we remove item j from the load, what do we know
about the remaining load?
 A: remainder must be the most valuable load
weighing at most W - wj that thief could take from
museum, excluding item j

David Luebke 18
12/09/21
 Solving The Knapsack Problem
 The optimal solution to the fractional knapsack
problem can be found with a greedy algorithm
 How?
 The optimal solution to the 0-1 problem cannot be
found with the same greedy strategy
 Greedy strategy: take in order of dollars/pound
 Example: 3 items weighing 10, 20, and 30 pounds,
knapsack can hold 50 pounds
 Suppose item 2 is worth $100. Assign values to the other items
so that the greedy strategy will fail

David Luebke 19
12/09/21
 The Knapsack Problem:
Greedy Vs. Dynamic
 The fractional problem can be solved greedily
 The 0-1 problem cannot be solved with a
greedy approach
 It can, however, be solved with dynamic
programming
 Homework 5: give a dynamic programming
solution to the 0-1 knapsack problem

David Luebke 20
12/09/21
 The End

David Luebke 21
12/09/21