Strong Induction with Fibonacci numbers

  • Thread starter Thread starter gwial
  • Start date Start date
  • Tags Tags
    Induction Numbers
Join the discussion
Ask a follow-up here, or get your own question answered by working scientists, mathematicians and engineers — people, not an autocomplete.
Real named experts · corrections over time · the nuance an AI answer skips
7 replies · 5K views
gwial
Messages
3
Reaction score
0

Homework Statement



Use strong mathematical induction to prove that the Fibonacci numbers satisfy the inequality fn > (√2)n

Homework Equations



for all integers n > 6. The Fibonacci numbers fn are defined recursively
by: f0 =0,f1 =1

For all n > 1, fn = fn−1 + fn−2

The Attempt at a Solution


My problem is, i really don't know where to start. I assuming that I am trying to prove that LHS>RHS in my base cases, instead of LHS=RHS. I started by attempting bases cases of 0 and 1. Those attempts really lead me nowhere. I then attempted 2 and 3 as my base cases. I got LHS<RHS for both, which based on my assumption is the inverse of what I am trying to accomplish.

Should i be looking at larger numbers for my base cases? "for all integers n >6" does this mean i should start bases at 7 and 8? And should i still use my assumption to prove LHS>RHS or should i be trying to make LHS=RHS anyways.

Thanks
 
Physics news on Phys.org
gwial said:
Should i be looking at larger numbers for my base cases? "for all integers n >6" does this mean i should start bases at 7 and 8?
Yes, it does, because the statement is not true when [tex]n \leq 6[/tex].
gwial said:
And should i still use my assumption to prove LHS>RHS or should i be trying to make LHS=RHS anyways.
You should be trying to use the inductive hypotheses [tex]f_{n-2} > \sqrt{2}(n - 2)[/tex] and [tex]f_{n-1} > \sqrt{2}(n-1)[/tex] to prove the next case of the statement, [tex]f_n > \sqrt{2}\,n[/tex].
 
Ok perfect thanks for the right direction.
 
It's exactly the same method as last proof, except we use base case of 7 and 8, then set up the same way as last proof we did on the previous assignment, you get down to f(n)> (root2)(n-1) + (root2)(n-2) and we are guaranteed that (root2)(n-2) is always less than (root2)(n-1) for n>6. So make the substitution where you have (root2)(n-1) with (root2)(n-2) to get f(n)> 2(root2)(n-2) which you can then write as f(n)> (root2)2 x (root2)(n-2) which you can see as it is like bases becomes (root2)(n-2+2) which simplifies to (root2)(n) . Qed. correct?
 
What you really NEED to do is state the problem correctly! As given, you cannot prove that "Fibonacci numbers satisfy the inequality fn > (√2)n" because it is NOT true. f0= 0 is NOT larger than [itex]0\sqrt{2}[/itex]. f1= 1 is NOT larger than [itex]1\sqrt{2}[/itex]. f2= 2 is NOT larger than [itex]2\sqrt{2}[/itex]. f3= 3 is NOT larger than [itex]3\sqrt{2}[/itex]. f4= 5 is NOT larger than [itex]4\sqrt{2}[/itex]. f5= 8 Is larger than [itex]5\sqrt{2}. Was the problem to prove that "Fibonacci numbers satisfy the inequality fn > (√2)n as long as n> 4"?[/itex]
 
No it should have said for n>6.
 
Says Polya in "How to solve it" - 'when you have the answer it isn't finished.'

Can you generalise your result? You would then see the question less from under the floorboards if you know what I mean, get more confidence.

This statement with √2 and 6 in it is almost misleadingly special. You could make a more general statement. You don't even have to work out anything involving particular numbers.
 
Ok, sorry if i put one line in section 2 when it was suppose to be in section 1. Thats just how our prof stated the question