A problem of mathematical induction

AI Thread Summary
The discussion centers on the necessity of proving the base case in mathematical induction, specifically why it is essential to establish the truth of a statement for n=1 before proceeding to the inductive step. Participants emphasize that without verifying the base case, the entire induction process is flawed, akin to ensuring the first domino falls in a domino effect. They illustrate this with examples, such as the sum of squares and the statement n^2 > √n, highlighting that failing to prove the base case can lead to incorrect conclusions for all natural numbers. The conversation also touches on the concept of starting induction from a different natural number, suggesting a more general approach to induction. Ultimately, the base case is crucial for the validity of the inductive argument.
sahilmm15
Messages
100
Reaction score
27
I have gone through the principle of mathematical induction. I cannot understand why do we need to prove every statement for n=1. I mean why is it necessary? Why can't we start directly from n=k then n=k+1. For example see the below image. Thanks!
 

Attachments

  • 16085543526946773401524287736839.jpg
    16085543526946773401524287736839.jpg
    46.1 KB · Views: 204
Physics news on Phys.org
If you can prove it for n = k you don't need induction
 
BvU said:
If you can prove it for n = k you don't need induction
Correct me if I am wrong. By seeing the 'domino' analogy we use the base case like p(1) or p(0) to actually 'check' whether the domino in the 'first' place has actually fallen or not. If not then there is no point in proving the statement. If it is true then we can carry with our process of proving the whole statement true.
 
sahilmm15 said:
I cannot understand why do we need to prove every statement for n=1
Induction goes like

IF ## \Biggl [ \ p(k) \Rightarrow p(k+1)\ \Biggr ] ## AND ##\ \ p(1)\ \ ## THEN ##\ \forall n: \ \ p(n) ##

so you don't prove ##p(k)## but only prove that IF p(k) THEN p(k+1)
 
sahilmm15 said:
Correct me if I am wrong. By seeing the 'domino' analogy we use the base case like p(1) or p(0) to actually 'check' whether the domino in the 'first' place has actually fallen or not. If not then there is no point in proving the statement. If it is true then we can carry with our process of proving the whole statement true.

This is right. So what's your question?

To see an example at work where the base case is important, take the formula for the sum of squares that's in your example, and add 1 to it. The inductive step will still work just fine.
 
BvU said:
Induction goes like

IF ## \Biggl [ \ p(k) \Rightarrow p(k+1)\ \Biggr ] ## AND ##\ \ p(1)\ \ ## THEN ##\ \forall n: \ \ p(n) ##

so you don't prove ##p(k)## but only prove that IF p(k) THEN p(k+1)
Thanks I am clear now. I like how this community is so active, they just answer instantaneously. It makes me easier to understand. Thanks again !
 
As a trivial example, consider the statement ##n = n + 1##.
Is really easy to prove ##p(k) \Longrightarrow p(k+1)## but that doesn't prove the above statement (which obviously is false), so you can see why we need to prove ##p(1)##.
 
  • Like
Likes Mark44, Klystron, fresh_42 and 1 other person
May i suggest another example like ##n^2>\sqrt{n}## For ##n=1## it is wrong, but trying to prove it without ##p(1)## leads to wrong results for every natural number. The more general statement of mathematical induction where we start from another natural number as a starting point not ##1## is needed i think.

It could be ##p(k)##, where ##k## is a natural number for that starting point.
 
Last edited by a moderator:
Well, if you have some statement ##F(n)## that is not true for ##n=1##, but you can prove ##F(k)\Longrightarrow F(k+1)## and you can prove ##F(n_0)## for some ##n_0##. Defining a "more general induction" isn't essentially useing induction over the statement
$$G(n) = F(n + n_0 - 1)$$
?

Concretelly, proving $$n^2 > \sqrt{n}, \quad \forall n>1$$
is equivalent to prove the statement ##(n+1)^2 > \sqrt{n+1}## by induction, which can be proved for ##n=1##
 
Back
Top