How to get regression plot for validation data using trainbr training function using MATLAB's nntool?

2 次查看（过去 30 天）
Sachin Bharadwaj M
Sachin Bharadwaj M 2021-6-24
编辑： Aneela 2024-4-23
In nntool, when trainbr algorithm is used, MATLAB is not providing Regression plot for validation data but only for train and test data. However, it is displaying Regression plot for training, validation & test data while using trainlm training algorithm. For reference the figures are shown below:
FIGURE SHOWING TRAINBR REGRESSION PLOT & OTHER MODEL DETAILS
FIGURE SHOWING TRAINLM REGRESSION PLOT & OTHER MODEL DETAILS
Please help me in debugging this issue. Thanks in advance!

请先登录，再回答此问题。

回答（1 个）

Aneela
Aneela 2024-4-23
编辑：Aneela 2024-4-23
Hi Sachin Bharadwaj,
I found below statement from the MathWorks Documentation.
The function “trainbr” that performs Bayesian Regularization disables validation stops by default.
For more information on “trainbr”, refer to the following link: https://www.mathworks.com/help/deeplearning/ref/trainbr.html#:~:text=Validation%20stops%20are,strictly%20positive%20value.
I also found an answer to a similar question in community, which states:
  • “max_fail” denotes the maximum number of times that we allow the validation to improve or to not improve before terminating training.
  • By default, max_fail is set to 0.
  • This can be the reason for MATLAB not providing regression plot for validation data.
Refer to the following answer by @MathWorks Support Team for a detailed explanation about how “trainbr” works:
https://www.mathworks.com/support/search.html/answers/405727-why-does-the-trainbr-function-not-require-a-validation-dataset.html

类别

在 Help Center 和 File Exchange 中查找有关 Surface and Mesh Plots 的更多信息

标签

产品


版本

R2020b

Community Treasure Hunt

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

Start Hunting!

Translated by