Waikato / wekaDeeplearning4j

Weka package for the Deeplearning4j java library
https://deeplearning.cms.waikato.ac.nz/
GNU General Public License v3.0
185 stars 202 forks source link

how to apply 1-d convolution? #55

Closed hahakid closed 4 years ago

hahakid commented 5 years ago

I tried to change the filter with 3*1 but the input is error

eibe commented 5 years ago

That should work. What was the exact error? Are you sure the stride parameter, etc., were consistent with the format of your input data?

hahakid commented 5 years ago

@eibe I will try it later, maybe caused by input error. feedforward input not conv input, something like that.

braun-steven commented 5 years ago

Keep in mind that you need a dataset iterator that supports convolutions. If you share the exact error with us we might be able to help you further.

Check out https://deeplearning.cms.waikato.ac.nz/user-guide/data/ for help on dataset iterators and https://deeplearning.cms.waikato.ac.nz/examples/classifying-mnist/ for an example on the MNIST dataset.

braun-steven commented 4 years ago

Since there is no further response, I will close this issue.