Train nonlinear input-output neural network with multiple data sets

3 次查看（过去 30 天）
Christopher Fernandez
回答： Greg Heath 2018-9-2
I have 5 sets of inputs corresponding outputs (all 1X100 data points) that I want to use to train a nonlinear input-output network. I'm trying to avoid combining inputs into one large set to avoid jumps in the data and because the network will be used with 1X100 size inputs. Is there a way to adjust the script that the GUI creates, or combine the data, to allow for a network to be trained multiple times with different data sets?

请先登录，再回答此问题。

回答（1 个）

Greg Heath
Greg Heath 2018-9-2
ALL OF YOUR FEARS ARE UNFOUNDED. IN PARTICULAR
1. Combining inputs into one large set to obtain jumps in the data can be beneficial
2. For regression and classification: As long as the training set is large and diverse enough to obtain relatively stable weight estimates and the test set is large enough to obtain relatively stable output estimates, the relative sizes of the training and operational data sets is irrelevant.
Hope this helps
*Thank you for formally accepting my answer*
Greg

类别

在 Help Center 和 File Exchange 中查找有关 Sequence and Numeric Feature Data Workflows 的更多信息

标签

产品


版本

R2017b

Community Treasure Hunt

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

Start Hunting!

Translated by