One day, Alice encountered this problem on an exam:
Given that , if has exactly two roots, then find the range of and prove it.
By using the derivative and its graph, Alice quickly reach the conclusion that .
But one thing troubled her greatly: Since she is a high school student and it's a standard exam, she is not permitted to use any definition of limit to prove the existence of roots.
"What a joke," she thought, but she quickly calmed down and came up with an idea: "You are not letting me use any definition of limit, but what I am really doing is using a simpler version of Epsilon-Delta Definition . If I could prove that for all , I can always find such that , then I can prove the existence of roots by Intermediate Value Theorem."
For the positive root, she quickly tried and , and she was very delighted.
For the negative root, she thought for a minute and decided to try , where is a constant. Now she was having trouble deciding what to choose.
Can you help her choose a proper , so that for all , ?
This section requires Javascript.
You are seeing this because something didn't load right. We suggest you, (a) try
refreshing the page, (b) enabling javascript if it is disabled on your browser and,
finally, (c)
loading the
non-javascript version of this page
. We're sorry about the hassle.
We have δ = − a n
So that f ( δ ) = ( − a n − 1 ) e − a n + a ( − a n ) 2
= − ( a n + 1 ) e − a n + a 2 n + 1
Now finding the value of n so that f ( δ ) > 0 for all a > 0
− ( a n + 1 ) e − a n + a 2 n + 1 > 0
⇒ a 2 n + 1 > ( a n + 1 ) e − a n
Both LHS and RHS are positive for any a > 0 and any n . So taking natural logarithm on both side.
( 2 n + 1 ) log a > log ( a n + 1 ) − a n ⋯ Eq. 1
Now considering RHS.
Let g ( x ) = log ( 1 + x ) − x
⇒ g ′ ( x ) = 1 + x 1 − 1 = 1 + x − x
We see that g ′ ( x ) = 0 at x = 0 and g ′ ( x ) < 0 for all x > 0 . Also, g ( 0 ) = 0
All these implies that g ( x ) < 0 for all x > 0 .
For any a > 0 , a n > 0 . So because of above argument, log ( a n + 1 ) − a n < 0 for all a > 0 .
For Eq. 1 to be true we have to choose n so that
( 2 n + 1 ) log a ≥ 0 for all a > 0
log a can take any real value for a > 0 . So the above inequation can only be true when coefficient of log a is 0 .
2 n + 1 = 0 ⇒ n = − 2 1