Seems like there are multiple bugs in the SGAN code.
I tried to fix them but I am not familiar with Keras yet so i can't fixed them all by my self.
Thanks in advance for any advice or update. :)
PS. Also, I would like to adapt to model to use with other data (person-based data, a row means a person with about 200 features/cols and output labeled as 0 or 1). It would be nice if you please clarify how to do that as well.
Seems like there are multiple bugs in the SGAN code. I tried to fix them but I am not familiar with Keras yet so i can't fixed them all by my self.
Thanks in advance for any advice or update. :)
PS. Also, I would like to adapt to model to use with other data (person-based data, a row means a person with about 200 features/cols and output labeled as 0 or 1). It would be nice if you please clarify how to do that as well.