How to evaluate the neural network by adjusted r-squared?

11 次查看（过去 30 天）
kokin
kokin 2015-11-22
编辑： Greg Heath 2015-11-25
Hi,everyone. I am using a back-propagation neural networks(BPNNs) to fit a economic nonlinear curve. The structure of the BPNNs is 7 nodes in input layer, 10 nodes in hidden layer and 1 node in output layer. Moreover, the dataset holds 36 samples. After training, I want to use the adjusted R-squared to evaluating the performance on the regressed curve of BPNNs. How can I define the n and p in the adjusted R-squared formula in here? And, is it right to use the adjusted R-squared to evaluating the neural networks? Thank you!!

请先登录，再回答此问题。

采纳的回答

Greg Heath
Greg Heath 2015-11-23
Search the NEWSGROUP and ANSWERS using
greg R2a
Hope this helps.
Thank you for formally accepting my answer
Greg
  1 个评论
Greg Heath
Greg Heath 2015-11-25
编辑：Greg Heath 2015-11-25
The only reason to evaluate a net using R2a is because you do not have enough data to use a sufficiently large nontraining test set.

请先登录，再进行评论。

更多回答（0 个）

类别

在 Help Center 和 File Exchange 中查找有关 Pattern Recognition and Classification 的更多信息

标签

Community Treasure Hunt

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

Start Hunting!

Translated by