Is my proof of the Fibonacci sequence correct?

264 Views Asked by At

Been working on this for some time now but have no idea if it's correct! Any hints are appreciated.

Recall the Fibonacci sequence: $f_1 = 1$, $f_2 = 1$, and for $n \geq 1$, $f_{n+2} = f_{n+1} + f_n$. Prove that $f_n > (\frac{5}{4})^n\ \forall \ n \geq 3$.

My answer:

"base case"

$[f_3 = 2 > (\frac{5}{4})^3\ = \frac{125}{64}\ correct$

$[f_4 = 3 > (\frac{5}{4})^4\ = \frac{625}{256}\ correct$

$assume\ f_k > (\frac{5}{4})^k\ for\ some\ k \geq 3$

$[and\ f_{k-1} > (\frac{5}{4})^{k-1}$

$then \ f_{k+1} = f_k + f_{k+1} > (\frac{5}{4})^k + (\frac{5}{4})^{k-1}$

$so \ f_{k+1} > (\frac{5}{4})^k + (\frac{5}{4})^{k-1}\ > (\frac{5}{4})^k (\frac{5}{4})^k = (\frac{5}{4})((\frac{5}{4})^k) = (\frac{5}{4})^{k+1}$

$f_{k+1} > (\frac{5}{4})^{k+1}$

$so \ f_n > (\frac{5}{4})^n \ \forall \ n \geq 3$

QED

2

There are 2 best solutions below

0
On BEST ANSWER

As J.W. Tanner mentioned, it's not true that $$\left(\frac{5}{4} \right)^k+ \left(\frac{5}{4} \right)^{k-1} > \left(\frac{5}{4} \right)^k\left(\frac{5}{4} \right)^k$$

(consider for example $k=3$ then $\frac{125}{64} + \frac{25}{16} < \frac{125} {64} \times\frac{125}{64}$

Hint: You can consider $$\left(\frac{5}{4} \right)^k+ \left(\frac{5}{4} \right)^{k-1} = \left(\frac{5}{4} \right)^{k-1} \left(\frac{5}{4} + 1\right) = \left(\frac{5}{4} \right)^{k-1} \left(\frac{9}{4} \right) > \left(\frac{5}{4} \right)^{k-1} \left(\frac{5}{4} \right)^{2}$$

and proceed from there.

0
On

You can use strong induction.

Start with $f_k = f_1 + f_2 + ... + f_{k-1} = 2 + ... + f_{k-1}$

Assume the proposition holds true for every $f_m$ where $m \leq k-1$.

Then we can bound $f_k$ using the geometric series sum as:

$\displaystyle f_k > 2 + \frac{(\frac 54)^3((\frac 54)^{k-3}-1)}{\frac 54 -1}$

and the expression on the right can be simplified into the form $A + B(\frac 54)^k$ (you're left to find constants $A$ and $B$).

Using the same method (and assuming the strong inductive hypothesis), you can quickly find the expression for $f_{k-1}$. It'll just be $A + \frac 45B(\frac 54)^k$.

Add those up (you're finding a lower bound on $f_{k+1}$) and show that this is greater than $(\frac 54)^{k+1} = \frac 54(\frac 54)^k$.

You will likely find it easier (I did) to put $(\frac 54)^k = x$, solve for the range of $x$ that satisfies the inequality and deduce that it the lower bound of such $x$ is lower than $\frac 54$, which naturally means it's lower than any value of $(\frac 54)^k$ for $k \geq 3$.

You've then established that $f_{k-1} + f_k = f_{k+1} > (\frac 54)^{k+1}$ and proven the proposition by strong induction. The base cases, you can handle.