b*******n 发帖数: 61 | 1 For a real n dimension function(derivative up to any order) if you only have one local minimizer it must be the global minimizer.
Is this true??
The function is defined in the whole n-dim real space, not in a interval.
thanks |
n******t 发帖数: 4406 | 2 如果没有别的条件显然不行了.
随便找个不连续就不成立了.
must
【在 b*******n 的大作中提到】 : For a real n dimension function(derivative up to any order) if you only have one local minimizer it must be the global minimizer. : Is this true?? : The function is defined in the whole n-dim real space, not in a interval. : thanks
|
b*******n 发帖数: 61 | 3 thanks. I changed it
【在 n******t 的大作中提到】 : 如果没有别的条件显然不行了. : 随便找个不连续就不成立了. : : must
|
l*****e 发帖数: 238 | 4 by minimizer, u mean minimum, right?
think of y=x(x-1)(x+1)
it has a unique local minimum between 0 and 1, but it's not a global minimum
have one local minimizer it must be the global minimizer.
【在 b*******n 的大作中提到】 : For a real n dimension function(derivative up to any order) if you only have one local minimizer it must be the global minimizer. : Is this true?? : The function is defined in the whole n-dim real space, not in a interval. : thanks
|