You are on page 1of 5

Case Study:

1) P=NP

If the solution to a problem is easy to check for correctness, is the problem easy to solve?

This is one of the most important unanswered questions in the field of computer science. It is also a
millennium prize problem which has a 1Milion$ prize for anyone who manages to prove or disprove the
validity of the given statement.

The problem refers to checking whether or not the difficulty of all types of problems scale up the same
way as we increase the size of the problem. For example, whether something like multiplication and
Sudoku both scale up in difficulty at the same rate as we increase the size of the numbers that were
multiplying or increase the size of the Sudoku grid.

The P stands for Polynomial Time and refers to how the difficulty of these classes of problems can all be
expressed as a polynomial function of its size. This class includes problems like multiplication, solving a
rubrics cube, finding out if a number is prime, etc.

These problems dont get too difficult too fast and through optimizing computer algorithms, we can
come up with a sufficiently fast program to solve these types of problems in polynomial time. In fact,
since our computer processors are becoming exponentially more efficient, any problem that falls under
this category will eventually become trivial to solve using computers. Example: How multiplication used
to be difficult with old computers but now our phones can easily multiply 20 digit numbers in less than a
second.

NP stands for nondeterministic polynomial time. NP refers to all classes of problems where we can
check if a particular answer is correct extremely fast (as a function of size in polynomial time). Therefore
all P problems also fall under NP. However, there are a huge amount of problems that exist in NP but
not in P. For example, Sudoku, Circuit design, Vehicle routing, Protein folding, etc.

In these examples, using algorithms to find a solution can take an enormous amount of time as the
difficulty is increased but we can easily verify a given solution. These problems have their difficulty being
expressed as a non-polynomial function of their size.

The entire problem is about whether or not all these different NP problems can eventually be reduced
to P type problems. The significance of this problem is that if it is proven true, that means that HUGE
problems such as protein folding (Which can be used to cure cancer) will eventually become trivial as
our computing power increases.

However most computer scientists have hypothesized that P != NP, but this hasnt stopped countless
scientists from trying to come up with a proof for the problem. If proven true, our world will never be
the same again.
2) Collatz Conjecture

Does the Collatz sequence eventually reach 1 for all positive integer initial values?

The Collatz conjecture is a conjecture in mathematics named after Lothar Collatz. This is often described
as one of the most difficult conjectures to prove and it is often said that mathematics as we know it,
isnt developed enough to tackle this problem.

The problem however is extremely easy to lay out and understand. We start with any number n. If n is
even, we divide it by 2. If n is odd, we multiply it by 3 and add 1 to it. Therefore, n/2 if n is even and
3n+1 if its odd.

The conjecture states that the following sequence will always terminate in 1 irrespective of the number
n. Let us run a particular sequence:

3105168421. As we can see, after we reach 1, the sequence begins to loop back so we
consider the sequence as terminated. Even using computer algorithms, it has been found that the
sequence terminates at 1 for every number tried up to date. Clearly there is some deep underlying
principle at work here which mathematicians havent been able to understand.
This is a sequence tree which shows the true structure of the sequence for various numbers n. It
appears random and chaotic but seems to indicate that there is some underlying principle which is
responsible for its branched state of existence.

The significance of finding a proof for the validity of the collatz conjecture is that the mathematics that
will be invented in an attempt to solve it will open huge avenues which in turn can be used to help crack
a variety of other problems. This is because it seems that our current system of mathematics isnt
enough to even begin a promising approach to this problem.
How to approach the idea of inventing a new type of mathematics for a proof:
Let us analyze the approach needed in order to invent mathematics by taking the case of an infinite
series. (1/2)+(1/4)+(1/8)+=1

If we as mathematicians wanted to prove this relation without knowing any of the underlying principles
behind infinite series, how would we go about it?

Our first approach would be to simply try adding the terms of this infinite series and seeing what
happens when we terminate the sequence at a finite point. We would quickly realize that this sum
would never actually explode to infinity and would actually be fairly well contained.

But how do we go about to make this assertion more rigorous? Well we could start by pondering the
nature of distance. Let us imagine the number line between the numbers 0 to 1. We are aware that
there are an infinite number of numbers between the two.

If we divide the line segment, we would end up with (1/2). But what now happens if we divide the
second half of the line segment by half again? We would hit (3/4) which happens to be a distance of
(1/4) away from the half-way point and so on. This thought exercise gives us a better understanding of
whats actually happening. Since we are able to constantly divide a line segment into multiple halves, we
can assume that the infinite sum of all these halves would end up giving us 1.

The numerical representation of that would be to equate the distances. We know that the distance
between 0 and 1 would be 1. We would also know that the sum of all the halves has to be =1 as we have
taken them to be between 0 and 1. So we can take the distances of each of those halves as
(1/2)+(1/4)+(1/8)+.

As the 2 distances are equal, we can say that the infinite series is =1. A way to make this statement
rigorous is to imagine a circle of any arbitrary distance around the number 1. Let the radius of this circle
fall as small as you want. Irrespective of how small it is, we will eventually find a line segment from our
infinite sum within the region enclosed.

But now lets try to make this result more general. In order to do that we first must try to figure out at
which step did we make an arbitrary assumption. We didnt have to break up the line segment between
0 and 1 by a half constantly. We could have also broken it up as (9/10)+(9/100)+(9/1000)+

From this we can infer that the general formula would be (1-p) + (1-p)p + (1-p)p^2 + (1-p)p^3 + .=1

As long as we take a number between 0 and 1, we have obtained the general formula for expressing
their infinite sum.

So to recap:

We play around with the idea until we have a better grasp of the underlying principle

We attempt to make the proof rigorous by transforming the problem back into numerical terms.

We then generalize the problem by taking any arbitrary decision we made and replacing it with a
variable.

You might also like