No classification layer in MATLAB Deep network designer
47 views (last 30 days)
Show older comments
Hi,
I am trying to build an image classification network in MATLAB. I am using the inbuilt Deep Network Designer app. But in the end of the neural network, there is a classification layer which is totally missing. I designed the last network in March and at that time the layer was present but now it is missing. Attaching the screenshots for reference. Also the tabs to select the image ibrary and the training options are missing from the top.
0 Comments
Answers (2)
Muskan
on 26 Jun 2024
As per the offical domentation from MATLAB R2024a, "classificationLayer" is not recommend to use. Please use the "trainnet" function instead. Refer to the following documentation of "classificationLayer" for a better understanding:
For more understanding on how to use "trainnet" function in Deep Network Designer, please refer to the following documentation:
0 Comments
Aditya
on 26 Jun 2024
In R2024a, Deep Network Designer updated to use "dlnetwork" objects by default.
This version of Deep Network Designer does not include training and therefore the "Training" and "Data" tabs are not present
Additionally, "dlnetwork" objects do not require an output layer and therefore, output layers are not in the layer library.
All of the differences between the two versions are in the following documentation page: https://www.mathworks.com/help/releases/R2024a/deeplearning/ref/deepnetworkdesigner-app.html#mw_c01f379c-a044-4c49-a2c1-13b71e3b15dd
Using "dlnetwork" objects from "trainnet" is recommended from R2024a onwards however, if it is not possible to transition to "dlnetwork" as recommended, you can still use the "DAGNetwork"/"SeriesNetwork" version of the app with the "-v1" flag as
>> deepNetworkDesigner("-v1")
0 Comments
See Also
Categories
Find more on Sequence and Numeric Feature Data Workflows in Help Center and File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!