Question: ( c ) In this problem, we will learn a classic use case of Hoeffding s inequality. Imagine we have an algorithm for solving some

(c) In this problem, we will learn a classic use case of Hoeffdings inequality. Imagine we have an algorithm
for solving some decision problem (e.g., Is there a cat in the picture?). Suppose that the algorithm makes a
randomized decision and returns the correct answer with probability 1
2+\delta , for some \delta >0(which is just a
bit better than a random guess). To improve the performance, we run the algorithm N times and apply the
majority voting. Please show that, for any in (0,1), the answer is correct with probability at least 1\epsi , as
long as N >=(1/2)\delta
2
ln(\epsi
1
).(Hint: For each i, define Xi to be a Bernoulli random variable for which Xi =1
when the algorithm return the correct answer at the i-th trial. Under majority voting, we know that if the final
answer is incorrect, then X1+ X2++ XN <= N/2. Use the negative part of Hoeffdings inequality. This
scheme is usually called boosting randomized algorithms.)

Step by Step Solution

There are 3 Steps involved in it

1 Expert Approved Answer
Step: 1 Unlock blur-text-image
Question Has Been Solved by an Expert!

Get step-by-step solutions from verified subject matter experts

Step: 2 Unlock
Step: 3 Unlock

Students Have Also Explored These Related Databases Questions!