Skip to content

Comment on The "Soft Maximum" functionparent

Comments

Trying to learn math from Wikipedia is like trying to learn English from the dictionary.

As far as the subject matter goes, a convex function can just be thought of as a function with a global maximum/minimum and no other local maximum/minimum. I.E. The derivative is zero in only one place which is the global max/min. These functions can then be easily optimized by taking their derivatives and finding the zero. It is a really nice way to optimize a problem.

...a convex function can just be thought of as a function with a global maximum/minimum and no other local maximum/minimum.

This is false. The convex function exp(x) has no global max or min. The function f(x,y)=x^2 has infinitely many global minima.

The nice thing about convex functions is that the set of places where they achieve a minimum is a convex set: the empty set for exp(x) and the line x=0 for f(x,y)=x^2.

No, in this case, a convex function is one where you can draw a straight line between any two points on its graph and not have the line intersect the graph at a third point.

You are not contradicting what fgimenez said, its the same condition stated differently.

edit: if two statements are not essentially same (and I am wrong), could you please explain it.

The logarithm function is convex by this definition but has no maximum/minimum point and never has 0 derivative.

I have about 5 years of informal/independent mathematical study and routinely use Wikipedia to crack fields open. I usually skim the intro then head straight to the references.

Once I have a vague idea of what it is, I drill down MathWorld using the new keywords.

AboutSource Built by g1lg1l

Hackerly is an independent reader for Hacker News, built on the public HN API. Not affiliated with Y Combinator.