# EVT and Fermats to prove f'(c)=0

1. Dec 7, 2011

### NWeid1

1. The problem statement, all variables and given/known data
If f is differentiable on the interval [a,b] and f'(a)<0<f'(b), prove that there is a c with a<c<b for which f'(c)=0.

2. Relevant equations

3. The attempt at a solution
Well, I first tried to use IVT but I was having a hard time to I talked to my prof. and he said to use extreme value theorem and fermats theorem. So, by EVT, I know there will be an aboslute maximum and absolute minimum on [a,b]. By f'(a)<0<f'(b) I know that it will be decreasing and increasing and therefore will have a local miniimum on (a,b). Fermats theorem then says that if f(c) is a local extremum, then c must be a critical number of f. Which means the function WILL have a c to make f'(c)=0. I know this intuitively but I'm having a hard time rigorously proving it.

File size:
6.9 KB
Views:
131
2. Dec 8, 2011

no one?

3. Dec 8, 2011

### Staff: Mentor

Please state the foregoing without "it". What you have written is very unclear. "It" could refer to f'(a), 0, or f'(b). I suspect that none of these is the antecedent.

4. Dec 8, 2011

### NWeid1

Well, I first tried to use IVT but I was having a hard time to I talked to my prof. and he said to use extreme value theorem and fermats theorem. So, by EVT, I know that f will have an aboslute maximum and absolute minimum on [a,b]. By f'(a)<0<f'(b) I know that f will be decreasing and increasing and therefore will have a local miniimum on (a,b). Fermats theorem then says that if f(c) is a local extremum, then c must be a critical number of f. Which means the function WILL have a c to make f'(c)=0. I know this intuitively but I'm having a hard time rigorously proving it.

5. Dec 8, 2011

### Staff: Mentor

Can you say this more precisely?
Where will f be decreasing?
Where will f be increasing?

If you haven't guessed, I'm trying to get you to make clear, unambiguous statements.

6. Dec 8, 2011

### NWeid1

f will be decreasing on (a,o) and increasing on (0,b)

7. Dec 8, 2011

### gb7nash

Not necessarily. We only know that the slope of f(x) at a is negative, and the slope of f(x) at b is positive. We don't know the exact intervals f is increasing or decreasing.

8. Dec 8, 2011

### Staff: Mentor

In addition to what gb7nash said, you are assuming here that a < 0 and b > 0. All you have is the interval [a, b]. There is no indication of whether 0 is in the interval, to the left of it, or to the right of it. It's just some arbitrary interval.