I have to prove that the function $F(x) = 8x^{4} - 8x^{2} + 1$ is chaotic. I would like to use the definition of a chaotic function which is:

Let $F$ be $F: V$ -> $V$.

1) Sensitive dependance on initial conditions: it exists $\delta > 0$ such that it exists $y \in N$ and $n \ge 0$ such that $|f_{n}(x) - f_{n}(y)| > \delta$ for any $x \in V$ and any neighborhood $N$ of $x$.

2) Topologically transitive: it exists $k > 0$ such that $f_{k}(J) \cap I \ne 0$ for any pair of open sets $J$ and $J \subset V$.

3) Periodic points are dense in $V$.

I wanted to try the following method:

Find $G$ and $\phi$ such that $F o \phi = \phi o G$, prove that G is chaotic and then as a matter of fact that F is chaotic.

Here, Use the definition of "topologically conjugate" to prove that $F(x)=4x^3-3x$ is chaotic., you can see the method for the function $F(x) = 4x^{3} - 3x$.

The problem I have is to find a function $G$.

Thanks for reading,


  • $\begingroup$ You may be familiar with the dictum that period three implies chaos. Is your Question about a proof using "the definition of a chaotic function" or about one using the method outlined in the last part of your post? $\endgroup$ – hardmath Jan 22 '15 at 12:42
  • $\begingroup$ My question is about having a hint on what is G so I can prove that G is chaotic using the definition of a chaotic function and hence F is chaotic :) $\endgroup$ – Dust009 Jan 22 '15 at 12:47
  • 1
    $\begingroup$ @hardmath The problem with applying such a general theorem is that it gives you very little information on the chaotic domain. Take $f(x)=3.83 x(1-x)$, for example. This function has an orbit of period 3 - an attractive orbit, in fact, so it's easy to find. As a result the dynamics look rather tame. The chaotic domain is just a Cantor set. The function in question, however, is chaotic on an entire interval so it's characteristics are quite different. $\endgroup$ – Mark McClure Jan 22 '15 at 13:18

Let's take a look at the two graphs you mention.

enter image description here

It looks to me like the first is conjugate to $\theta \mapsto 4\theta \mod 2\pi$ while the second is conjugate to $\theta \mapsto 3\theta \mod 2\pi$. In fact, the conjugacy functions are both given by $\varphi(\theta) = \cos(n\theta)$ for $n=3,4$. This is because these are both Chebyshev polynomials and, in that context, their defining characteristic is that $T_n(\cos(\theta))=\cos(n\theta))$. Of course, $\theta \mapsto n\theta \mod 2\pi$ is well known to be chaotic.

  • 1
    $\begingroup$ Thank you, I think I get the point :) $\endgroup$ – Dust009 Jan 22 '15 at 13:18
  • $\begingroup$ nice 1 (+1-1+1-1+1), :P $\endgroup$ – Shobhit Jan 22 '15 at 13:18
  • $\begingroup$ @Dust009 Good, I'm glad you get it! If you have any further trouble, any exposition of the Chebyshev polynomials should explain the conjugacy. $\endgroup$ – Mark McClure Jan 22 '15 at 13:20

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.