InstructorNeelima Gupta
Expected Running Times and Randomized Algorithms
Instructor Neelima Gupta [email protected]
Expected Running Time of Insertion Sort
x1,x2,........., xi-1,xi,.......…,xn
For I = 2 to nInsert the ith element xi in the partially sorted
list x1,x2,........., xi-1.
(at rth position)
Expected Running Time of Insertion SortLet Xi be the random variable which represents the
number of comparisons required to insert ith element of the input array in the sorted sub array of first i-1 elements.
Xi : can take values 1…i-1 (denoted by
xi1,xi2,..................…,xii)
E(Xi) = Σj xijp(xij )where E(Xi) is the expected value Xi
And, p(xij) is the probability of inserting xi in the jth position 1≤j≤i
Expected Running Time of Insertion Sort
x1,x2,........., xi-1,xi,.......…,xn
How many comparisons it makes to insert ith element in jth position?
(at jth position)
Position # of Comparisionsi 1i-1 2i-2 3
. . . . . .
2 i-11 i-1
Note: Here, both position 2 and 1 have # of Comparisions equal to i-1. Why? Because to insert element at position 2 we have to compare with previously
first element. and after that comparison we know which of them come first and which at second.
Thus, E(Xi) = (1/i){ i-1Σk=1k + (i-1) }where 1/i is the probability to insert at jth position in the i possible positions.
For n elements,
E(X1 + X2 + .............+Xn) = nΣi=2 E(Xi)
= nΣi=2 (1/i){ i-1Σk=1k + (i-1) } = (n-1)(n-4)/4
Therefore average case of insertion sort takes Θ(n2)
For n number of elements, expected time taken is,
T = nΣi=2 (1/i){ i-1Σk=1k + (i-1) }where 1/i is the probability to insert at rth
position in the i possible positions.
E(X1 + X2 + .............+Xn) = nΣi=1 E(Xi)Where,Xi is expected value of inserting Xi element.
T = (n-1)(n-4)/4Therefore average case of insertion sort takes
Θ(n2)
Quick-Sort
Pick the first item from the array--call it the pivot Partition the items in the array around the pivot so all elements
to the left are to the pivot and all elements to the right are greater than the pivot
Use recursion to sort the two partitions
pivotpartition: items > pivotpartition 1: items pivot
Quicksort: Expected number of comparisons
Partition may generate splits (0:n-1, 1:n-2, 2:n-3, … , n-2:1, n-1:0)
each with probability 1/nIf T(n) is the expected running time,
Randomized Quick-Sort
Pick an element from the array--call it the pivot Partition the items in the array around the pivot so all elements
to the left are to the pivot and all elements to the right are greater than the pivot
Use recursion to sort the two partitions
pivotpartition: items > pivotpartition 1: items pivot
RemarksNot much different from the Q-sort except that
earlier, the algorithm was deterministic and the bounds were probabilistic.
Here the algorithm is also randomized. We pick an element to be a pivot randomly. Notice that there isn’t any difference as to how does the algorithm behave there onwards?
In the earlier case, we can identify the worst case input. Here no input is worst case.
Randomized Select
1
0
}1,max{1 n
k
nknTkTn
nT
Randomized AlgorithmsA randomized algorithm performs coin tosses (i.e.,
uses random bits) to control its execution
i ← random()if i = 0do A …else { i.e. i = 1}do B …
Its running time depends on the outcomes of the coin tosses
Assumptions
coins are unbiased, andcoin tosses are independent
The worst-case running time of a randomized algorithm may be large but occurs with very low probability (e.g., it occurs when all the coin tosses give “heads”)
Monte Carlo AlgorithmsRunning times are guaranteed but the output may
not be completely correct.
Probability of error is low.
Las Vegas AlgorithmsOutput is guaranteed to be correct.
Bounds on running times hold with high probability.
What type of algorithm is Randomized Qsort?
Why expected running times?Markov’s inequalityP( X > k E(X)) < 1/ki.e. the probability that the algorithm will take more
than O(2 E(X)) time is less than 1/2.Or the probability that the algorithm will take more than
O(10 E(X)) time is less than 1/10.This is the reason why Qsort does well in practice.
Markov’s Bound
P(X<kM)< 1/k , where k is a constant.
Chernouff’s Bound
P(X>2μ)< ½
A More Stronger Result
P(X>k μ )< 1/nk, where k is a constant.
Binary Search Tree
What is a binary search tree?
A BST is a possibly empty rooted tree with a key value, a possible empty left subtree and a possible empty right subtree.
Each of the left subtree and the right subtree is a BST.
Binary Search Tree
Pick the first item from the array--call it the pivot…it becomes the root of the BST.
Partition the items in the array around the pivot so that all elements to the left are the pivot and all elements to the right are greater than the pivot
Recursively Build a BST on each partition. They become the left and the right sub-tree of the root.
Binary Search TreeConsider the following input:
1,2,3 …………………10,000.
What is the time for construction?Search Time?
Randomly Built Binary Search Tree
Pick an item from the array randomly --call it the pivot…it becomes the root of the BST.
Partition the items in the array around the pivot so that all elements to the left are the pivot and all elements to the right are greater than the pivot
Recursively Build a BST on each partition. They become the left and the right sub-tree of the root.
ExampleConsider the input
10, 20, 30, 40, 50, 60, 70, 80, 90, 100.
WLOG, assume that the keys are distinct. (What if they are not?) Rank(x) = number of elements < xLet Xi : height of the tree rooted at a node with rank=i.Let Yi : exponential height of the tree=2^Xi Let H : height of the entire BST, then
H=max{H1,H2} + 1where H1 : ht. of left subtree H2 : ht.of right subtree
Height of the RBST
Y=2^H =2.max{2^H1,2^H2}
E(EH(T(n))): Expected value of exponential ht. of the tree with ‘n’ nodes.
E(EH(T(n)))=2/n ∑ max{EH(T(k)),EH(T(n-1-k))}=O(n^3)
E(H(T(n))) =E(log (EH(T(n)))) = O(log n)
Construction Time?Search Time?What is the worst case input?
AcknowledgementsKunal VermaNidhi Aggarwal
And other students of MSc(CS) batch 2009.
HashingMotivation: symbol tables
A compiler uses a symbol table to relate symbols to associated data Symbols: variable names, procedure names, etc. Associated data: memory location, call graph, etc.
For a symbol table (also called a dictionary), we care about search, insertion, and deletion
We typically don’t care about sorted order
Hash TablesMore formally:
Given a table T and a record x, with key (= symbol) and satellite data, we need to support: Insert (T, x) Delete (T, x) Search(T, x)
We want these to be fast, but don’t care about sorting the records
The structure we will use is a hash tableSupports all the above in O(1) expected time!
Hash FunctionsNext problem: collision T
0
m - 1
h(k1)h(k4)
h(k2) = h(k5)
h(k3)
k4
k2 k3
k1
k5
U(universe of keys)
K(actualkeys)
Resolving CollisionsHow can we solve the problem of collisions?One of the solution is : chainingOther solutions: open addressing
ChainingChaining puts elements that hash to the same slot in
a linked list:
——
——
——————
——T
k4
k2k3
k1
k5
U(universe of keys)
K(actualkeys)
k6k8
k7
k1 k4 ——
k5 k2
k3
k8 k6 ————
k7 ——
ChainingHow do we insert an element?
——
——
——————
——T
k4
k2k3
k1
k5
U(universe of keys)
K(actualkeys)
k6k8
k7
k1 k4 ——
k5 k2
k3
k8 k6 ————
k7 ——
ChainingHow do we delete an element?
——
——
——————
——T
k4
k2k3
k1
k5
U(universe of keys)
K(actualkeys)
k6k8
k7
k1 k4 ——
k5 k2
k3
k8 k6 ————
k7 ——
ChainingHow do we search for a element with a
given key?
——
——
——————
——T
k4
k2k3
k1
k5
U(universe of keys)
K(actualkeys)
k6k8
k7
k1 k4 ——
k5 k2
k3
k8 k6 ————
k7 ——
Analysis of ChainingAssume simple uniform hashing: each key in table is
equally likely to be hashed to any slotGiven n keys and m slots in the table: the
load factor = n/m = average # keys per slotWhat will be the average cost of an unsuccessful
search for a key?
Analysis of ChainingAssume simple uniform hashing: each key in table is
equally likely to be hashed to any slotGiven n keys and m slots in the table, the
load factor = n/m = average # keys per slotWhat will be the average cost of an unsuccessful
search for a key? A: O(1+)
Analysis of ChainingAssume simple uniform hashing: each key in
table is equally likely to be hashed to any slotGiven n keys and m slots in the table, the
load factor = n/m = average # keys per slotWhat will be the average cost of an
unsuccessful search for a key? A: O(1+)What will be the average cost of a successful
search?
Analysis of ChainingAssume simple uniform hashing: each key in
table is equally likely to be hashed to any slotGiven n keys and m slots in the table, the
load factor = n/m = average # keys per slotWhat will be the average cost of an
unsuccessful search for a key? A: O(1+)What will be the average cost of a successful
search? A: O((1 + )/2) = O(1 + )
Analysis of Chaining ContinuedSo the cost of searching = O(1 + )If the number of keys n is proportional to the number
of slots in the table, what is ? A: = O(1)
In other words, we can make the expected cost of searching constant if we make constant
If we could prove this,
P(failure)<1/k (we are sort of happy)
P(failure)<1/nk (most of times this is true and we’re
happy )
P(failure)<1/2n (this is difficult but still we want this)
A Final Word About Randomized Algorithms
AcknowledgementsKunal VermaNidhi Aggarwal
And other students of MSc(CS) batch 2009.
END