0
$\begingroup$

I'm trying to prove $f_n \leq \left(\frac{1 + \sqrt{5}}{2}\right)^{n-1}$ with induction, and I'm stuck in the induction step.

Basis: n = 2

$f_{2} \leq \left(\frac{1 + \sqrt{5}}{2}\right)^{2-1} \Rightarrow f_{2} \leq \left(\frac{1 + \sqrt{5}}{2}\right)^{1} \Rightarrow 1 \leq \approx 1.68$ Base case is true

Ind Hyp.

Prove $n= k$ holds for 2 up to k

Ind Step

Consider $k+1$, let $\varphi = \left(\frac{1 + \sqrt{5}}{2}\right)$

LHS: $f_{k+1}=f_{k} + f_{k-1}$

$f_{k} \leq \varphi^{k-1}$ and $f_{k-1} \leq \varphi^{k-2}$

$f_{k+1} = \frac{\varphi^{k}}{\varphi} + \frac{\varphi}{\varphi^{2}}$

$f_{k+1} = \varphi^{k}\left(\frac{1}{\varphi} + \frac{1}{\varphi^{2}}\right)$

Here is where I'm stuck, I'm not sure if this was the right direction to take, but I believe I need to prove the following to continue:

$f_{k+1}< \varphi^{(k+1)-1}$

Could some one shed some light on my problem or show me a direction I can take?

$\endgroup$
0
$\begingroup$

From $f_k\le\varphi^{k-1}$ and $f_{k-1}\le\varphi^{k-2}$ you get $f_{k+1}\le\varphi^{k-1}+\varphi^{k-2}$, not $f_{k+1}=\varphi^{k-1}+\varphi^{k-2}$. Then

$$f_{k+1}\le\varphi^{k-1}+\varphi^{k-2}=\varphi^k\left(\frac1\varphi+\frac1{\varphi^2}\right)=\varphi^k\left(\frac{\varphi+1}{\varphi^2}\right)\;,$$

and you’re done as soon as you show that $\varphi+1=\varphi^2$. You can do this either by direct computation or by using the fact that $\varphi$ is a solution of the quadratic $x^2-x-1=0$.

| cite | improve this answer | |
$\endgroup$

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.