s*******d
发帖数: 126 | 1 我的搜索法实际是在做测量，搜索过程中的每个实测样本点所对应的临界
模型不过是为了对该点作为可能的临界点时所具有的合并预测残差做出测量，以便作为
权重计算的依据。有了这个权重，又有了临界点的可测空间和实测样本点，就可以计算
临界点的加权期望。从而决定临界模型。事情就是这么简单。真的非常非常简单。
I learnt statistics in english, and I cannot, for the life of me, understand
this paragraph, even after I read it like seven times. If the idea is
simple, I'd like to know. A rough translation would do, doesn't have to be
very precise. I have a ph.d. in statistics, so I should be able to follow
the idea (I hope). Thanks. | T*******I
发帖数: 5138 | 2 Let me try.
It says:
My searching method is actually a process of measuring , in which a
threshold model corresponding to a real measured sample point that is
assumed as a possible threshold in the process is just used as a tool to
measure or explicit an expected combined absolute residuals of it, so that
it can be used to calculate a weight of the real sample point assumed as the
possible threhsold. Once we have the weight, a measurable space and the
real sample points, we can calculate a weighted expectation of the unknown
threshold, thus to determine the threshold model or piecewise model. The
method is simple like this. Really simple.
Wish it would be helpful.
understand
【在 s*******d 的大作中提到】
: 我的搜索法实际是在做测量，搜索过程中的每个实测样本点所对应的临界
: 模型不过是为了对该点作为可能的临界点时所具有的合并预测残差做出测量，以便作为
: 权重计算的依据。有了这个权重，又有了临界点的可测空间和实测样本点，就可以计算
: 临界点的加权期望。从而决定临界模型。事情就是这么简单。真的非常非常简单。
: I learnt statistics in english, and I cannot, for the life of me, understand
: this paragraph, even after I read it like seven times. If the idea is
: simple, I'd like to know. A rough translation would do, doesn't have to be
: very precise. I have a ph.d. in statistics, so I should be able to follow
: the idea (I hope). Thanks.
|
|
