Scribd
Upload
5 views

Lecture 6

Uploaded by

Free Coupon
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
5 views

Lecture 6

Uploaded by

Free Coupon
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/ 15

The Greedy Design

Technique-2
Another Greedy Design Problem: Fractional Knapsack
• Suppose that you are going for a long hike. You want to take some food
with you but the bag that you carry, only has a limited capacity.
• You aim is to fill your bag with as much expensive food item as possible.
• The problem is called fractional because you can cut out a portion of your
food item to put in your bag(knapsack).
• This optimization problem can be solved using greedy approach because
we are allowed to take a portion of our food items rather than taking it as a
whole.
• Each food item has a price associated with it.
The Fractional Knapsack Problem
Formalization
• Input: weights 𝑤1 … 𝑤𝑛 and values 𝑣1 … 𝑣𝑛 of 𝑛 items. Capacity 𝑊.
• Output: The maximum total value of fractions of items that fits into
the 𝑊.
Safe Move
• There exists an optimal solution that uses as much as possible of an
item with the maximal value per unit of weight.
The Greedy Algorithm
• While knapsack is not full.
𝑣𝑖
• Choose item 𝑖 with maximum .
𝑤𝑖
• If the item fits into the knapsack, take all of it.
• Otherwise take so much as to fill the knapsack.
• Return total value and amounts taken.
An Algorithm
1. Knapsack(𝑊, 𝑤, 𝑣)
2. let 𝐴[1 … 𝑤. 𝑙𝑒𝑛𝑔𝑡ℎ] be an array initialized to 0
3. 𝑉 = 0
𝑣[𝑖]
4. create an array of:
𝑤[𝑖]
5. for 𝑖 = 1 to 𝑤. 𝑙𝑒𝑛𝑔𝑡ℎ
6. if 𝑊 == 0
7. return (𝑉, 𝐴)
8. 𝑎 = min 𝑤[𝑖], 𝑊
𝑣[𝑖]
9. 𝑉 =𝑉+𝑎
𝑤[𝑖]
10. 𝑤[𝑖] = 𝑤[𝑖] − 𝑎
11. 𝐴[𝑖] = 𝑎
12. 𝑊 =𝑊−𝑎
13. return (𝑉, 𝐴)
Better one:
1. Knapsack(𝑊, 𝑤, 𝑣)
2. let 𝐴[1 … 𝑤. 𝑙𝑒𝑛𝑔𝑡ℎ] be an array initialized to 0
3. 𝑉 = 0
𝑣[𝑖]
4. //sort items by decreasing order of
𝑤[𝑖]
5. for 𝑖 = 1 to 𝑤. 𝑙𝑒𝑛𝑔𝑡ℎ
6. if 𝑊 == 0
7. return (𝑉, 𝐴)
8. 𝑎 = min 𝑤[𝑖], 𝑊
𝑣[𝑖]
9. 𝑉 =𝑉+𝑎
𝑤[𝑖]
10. 𝑤[𝑖] = 𝑤[𝑖] − 𝑎
11. 𝐴[𝑖] = 𝑎
12. 𝑊 =𝑊−𝑎
13. return (𝑉, 𝐴)
Some Observations
𝑣𝑖
• If we sort the input with respect to their for all 𝑖 then we will start
𝑤𝑖
with the item that has this ratio, the largest.
𝑣𝑖
• Otherwise we will have to repeatedly chose the item whose is
𝑤𝑖
higher than what has been seen before previously, costing us 𝑂 𝑛2
time.
The Running Time
• Note the runtime 𝑇 𝑛 = 𝑂 𝑛 + 𝑂(𝑛 lg 𝑛).
• The first 𝑂 𝑛 is the number of iterations and the second 𝑂(𝑛 lg 𝑛) is
the time taken for sorting.
• Since 𝑛 = 𝑂(𝑛 lg 𝑛) we can conclude that 𝑇 𝑛 = 𝑂(𝑛 lg 𝑛).
Review of The Greedy Strategy: Reduction to
Subproblem
• Make a first move
• Then solve a problem of the same kind
• Smaller: fewer weight capacity left, fewer fuel stations etc. This is
called a “subproblem”
Review of The Greedy Strategy: Safe Move
• A move is called safe if there is an optimal solution consistent with
this first move
• Not all first moves are safe

You might also like