MHB Solving Exercise - Proving f(1 + w) > f(1) + w*f(1)

  • Thread starter Thread starter IlanSherer
  • Start date Start date
  • Tags Tags
    Exercise
Click For Summary
The discussion focuses on proving the inequality f(1 + w) > f(1) + w*f(1) for a convex function f(x) that passes through the origin and remains in the first quadrant. It is established that a convex function's graph lies above the line segment connecting any two points on it, specifically between (0,0) and (1,f(1)). The argument is supported by demonstrating that for w > 0, the function's strict convexity ensures that f(1 + w) exceeds the linear approximation f(1) + w*f(1). The conclusion emphasizes that the strict convexity of f(x) is crucial for the inequality to hold true for all positive w. This mathematical proof highlights the properties of convex functions in relation to linear bounds.
IlanSherer
Messages
1
Reaction score
0
Hello :)

Help please about the following exercise :

The function f(x) comes up and is an convex which gets out from (0,0) and continues into the first quarter.
Prove that for all w > 0, takes place f(1 + w) > f(1) + w*f(1).

I have been trying for two days to solve it :\

Thanks a lot!
 
Physics news on Phys.org
IlanSherer said:
Hello :)

Help please about the following exercise :

The function f(x) comes up and is an convex which gets out from (0,0) and continues into the first quarter.
Prove that for all w > 0, takes place f(1 + w) > f(1) + w*f(1).

I have been trying for two days to solve it :\

Thanks a lot!

Hi IlanSherer! Welcome to MHB! (Smile)

A function is defined to be convex if the line segment between any two points on the graph of the function lies above or on the graph.

Let's pick the points $(0,0)$ and $(1,f(1))$.
Then any point on the corresponding line segment is above the graph, and any point on the line to the right, is below the graph.
And that line is given by $y=f(1) x$.
Yes?

It looks like this:
\begin{tikzpicture}[xscale=3]
\def\x{3};
\draw[<->] (0,0) -- (3.3,0) node
{$x$};
\draw[<->] (0,0) -- (0,9.5) node[above] {$y$};
\draw[help lines] (1,-0.1) node[below] {1} -- (1,1) -- (-0.03,1) node
{f(1)};
\draw[help lines] (\x,\x) -- (-0.03,\x) node
{f(1)+wf(1)};
\draw[help lines] (\x,-0.1) node[below] {1+w} -- (\x,{\x^2}) -- (-0.03,{\x^2}) node
{f(1+w)};
\draw[domain=-0:3.1, variable=\x, red, ultra thick] plot ({\x}, {(\x)^2}) node
{y=f(x)};
\draw[blue, ultra thick] (0,0) -- (3.2,3.2) node
{y=f(1)x};
\draw[blue, ultra thick] (0,0) -- (1,0) -- node
{f(1)} (1,1) -- node[above] {w} (\x,1) -- node
{wf(1)} (\x,\x);
\end{tikzpicture}

So for $w \ge 0$, we have:
$$f(1+w) \ge f(1)(1+w) = f(1) + wf(1)$$

And if the function is strictly convex, we have for $w>0$:
$$f(1+w) > f(1) + wf(1)$$​