Proving Convexity of $f: \mathbb{R} \to \mathbb{R}$

  • MHB
  • Thread starter evinda
  • Start date
  • Tags
    Convex
Excited)In summary, we are given a continuous function $f: \mathbb{R} \to \mathbb{R}$ and want to show that it is convex. We can use the given property of $f$ to show that $f(\lambda x + (1-\lambda)y) \leq \frac{f(\lambda x) + f( (1-\lambda)y)}{2}$, and then apply the definition of convexity to complete the proof.
  • #1
evinda
Gold Member
MHB
3,836
0
Hello! (Wave)

We are given a continuous function $f: \mathbb{R} \to \mathbb{R}$ such that $f \left( \frac{x+y}{2}\right) \leq \frac{f(x)+f(y)}{2}, \forall x, y \in \mathbb{R}$. I want to show that $f$ is convex.I have tried the following:

Let $\lambda \in [0,1]$.

We have that $f(\lambda x+(1-\lambda)y)=f \left( \frac{2 \lambda x+ 2(1-\lambda)y}{2}\right) \leq \frac{f(2 \lambda x)+f(2(1-\lambda)y)}{2}$.But can we show that the latter is $\leq \lambda f(x)+(1-\lambda)f(y)$ ?Or can't we show it by showing that the definition of convexity is satisfied?

If not, how else can we prove the desired result? (Thinking)
 
Physics news on Phys.org
  • #2
evinda said:
Hello! (Wave)

We are given a continuous function $f: \mathbb{R} \to \mathbb{R}$ such that $f \left( \frac{x+y}{2}\right) \leq \frac{f(x)+f(y)}{2}, \forall x, y \in \mathbb{R}$. I want to show that $f$ is convex.I have tried the following:

Let $\lambda \in [0,1]$.

We have that $f(\lambda x+(1-\lambda)y)=f \left( \frac{2 \lambda x+ 2(1-\lambda)y}{2}\right) \leq \frac{f(2 \lambda x)+f(2(1-\lambda)y)}{2}$.But can we show that the latter is $\leq \lambda f(x)+(1-\lambda)f(y)$ ?Or can't we show it by showing that the definition of convexity is satisfied?

If not, how else can we prove the desired result? (Thinking)
There does not seem to be any short proof of this result.

As a first step, $f\bigl(\frac14x + \frac34y\bigr) = f\bigl(\frac12\bigl(\frac12(x+y) + y\bigr)\bigr) \leqslant \frac12\bigl(f\bigl(\frac12(x+y)\bigr) + f(y)\bigr) \leqslant \frac12\bigl(\frac12\bigl(f(x) + f(y)\bigr) + f(y)\bigr) = \frac14f(x) + \frac34f(y).$

So the result holds for $\lambda = \frac14$. Now use that same argument to show (by induction on $n$) that the result holds when $\lambda = \dfrac m{2^n}$ ($m=1,2,\ldots, 2^n$).

Finally, use the continuity of $f$ to show that the result holds for all $\lambda \in [0,1]$ (because each such $\lambda$ can be approximated by dyadic rationals). That last step is essential, because it is known that there are discontinuous functions for which the result is false.
 
  • #3


Hi there! Interesting problem. Let's see if we can work through this together.

First, let's recall the definition of convexity. A function $f$ is convex if for any two points $x$ and $y$ in its domain and any $\lambda \in [0,1]$, we have $f(\lambda x + (1-\lambda)y) \leq \lambda f(x) + (1-\lambda)f(y)$.

Now, let's take a closer look at the expression we have: $f(\lambda x + (1-\lambda)y) \leq \frac{f(2\lambda x) + f(2(1-\lambda)y)}{2}$. Notice that the only difference between this expression and the definition of convexity is that we have $\frac{1}{2}$ on the right hand side.

But, we can use the given property of $f$ to our advantage. We can rewrite $f(2\lambda x)$ as $f(\lambda x + \lambda x)$, and similarly for $f(2(1-\lambda)y)$. Then, by the given property, we have $f(\lambda x + \lambda x) \leq \frac{f(\lambda x) + f(\lambda x)}{2} = f(\lambda x)$, and similarly for $f(2(1-\lambda)y)$. So, our expression becomes $f(\lambda x + (1-\lambda)y) \leq \frac{f(\lambda x) + f( (1-\lambda)y)}{2}$.

Now, all we have to do is apply the definition of convexity, and we get $f(\lambda x + (1-\lambda)y) \leq \lambda f(x) + (1-\lambda)f(y)$, which is exactly what we wanted to show.

I hope this helps! Let me know if you have any other questions.
 

Related to Proving Convexity of $f: \mathbb{R} \to \mathbb{R}$

What does it mean for a function to be convex?

Convexity is a property of a function where the line segment connecting any two points on the graph of the function always lies above or on the graph. In other words, a function is convex if it is always "curving up" and has no "dips" or "valleys."

How do you prove that a function is convex?

To prove that a function is convex, we must show that for any two points on the graph, the line segment connecting them lies above or on the graph. This can be done by using the definition of convexity and algebraic manipulation.

What are some common examples of convex functions?

Linear functions, quadratic functions, and exponential functions are all examples of convex functions. Additionally, functions that are increasing and have a positive second derivative are also convex.

Can a function be both convex and concave?

No, a function cannot be both convex and concave. These are opposite properties of a function and a function can only have one or the other.

Are all convex functions also continuous?

No, not all convex functions are continuous. A function can be convex but still have discontinuities or sharp corners. However, all continuous convex functions are also differentiable.

Similar threads

  • Topology and Analysis
Replies
24
Views
2K
  • Topology and Analysis
Replies
16
Views
1K
Replies
4
Views
398
Replies
6
Views
888
Replies
1
Views
228
  • Topology and Analysis
Replies
9
Views
2K
  • Topology and Analysis
Replies
2
Views
190
Replies
4
Views
779
  • Topology and Analysis
2
Replies
38
Views
3K
Replies
1
Views
1K
Back
Top