How can I extrapolate data using regression learner app?

I have great solutions using the regression learner app. If I export the model and use the same data, then the result differs a lot from the solution in the regression app. I have already read the documentation, but this does not help. Is there someone how is more familiar with this methodology?
Unbenannt.PNG

回答 (2 件)

Star Strider
Star Strider 2018 年 12 月 19 日

1 投票

You have not posted what you are doing or the model you are using, so I suggest that you click on Export Model, then post the exported code for your model here.
Extrapolating to a small extent can work for some models, although for others (such as polynomial models), it can produce inappropriate results. As a general rule, it is best not to extrapolate beyond the region of fit.
shahar katz
shahar katz 2018 年 12 月 19 日

0 投票

after you export the modle to - "trainedModel", you can predict simply by -
yfit = trainedModel.predictFcn(data)
For 1d simple interpolation i would suggest interp1 with extrap option. becuase with this regression you need to have a better understanding of what you are doing.

カテゴリ

質問済み:

2018 年 12 月 19 日

回答済み:

2018 年 12 月 19 日

Community Treasure Hunt

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

Start Hunting!

Translated by