Curve fitting toolbox define max error

V
V
2020 7 5
0 个回答
更新时间：2020 7 5
4 次查看（30 天）

0 个投票

I am using curve fitting toolbox for Z=F(x,y)
I am getting results in this below form. Where I can do the setting where I can define max allowed error 0.09 .i.e max I put x & y backin generated equation max devaition I should get from Z is +-0.09 only. This is max/min error in "Z" where I can define that.
Goodness of fit:
SSE: 1.08e+07
R-square: 0.994
Adjusted R-square: 0.994
RMSE: 16.8

3 个评论
显示 1更早的评论 隐藏 1更早的评论

John D'Errico
John D'Errico 2020-7-5
Are you confused as to what maximum error might mean? Or, as you asking how to set such a number as a maximum?
The curve fitting toolbox does not provide an option to let you indicate what the maximum residual for any data point will be, thus Z +/- some tolerance.. Surely you should see that you can always choose a model that fits so poorly that this could never be a viable option.
V
V 2020-7-5
Will polynomial fit or any other fit equation is formed, it will give a equation, in which when x and y are fitted, will return Zeq.
I consider error = Zer-Zoriginal
the absolute error should not increase +-0.09999.
This is what I am aksing, is there any setting to do that?
dpb
dpb 2020-7-5
As John, says, "No".
It isn't a necessarily fulfillable condition for a specific model (order of polynomial, say).
You could try successive orders of a polynomial until met the condition altho would be easily prone to overfitting that way.

请先登录，再进行评论。

请先登录，再回答此问题。

回答（0 个）

类别

在 帮助中心 和 File Exchange 中查找有关 Linear and Nonlinear Regression 的更多信息

标签

提问:

V
V
2020-7-5

评论：

dpb
dpb
2020-7-5

Community Treasure Hunt

Find the treasures in MATLAB Central and discover how the community can help you!

Start Hunting!

Translated by