Does dta works with multi-label data or continious target variable?
Closed this issue · 6 comments
Hi, does dta supports non-binary target input?
In my case I am trying to generate continuous features (which is my future target for regression purposes) of multi-label dataset.
The run.recorder.plot() gives me this.
dta supports continuous data as input variables: https://lschmiddey.github.io/fastpages_/2021/11/07/DeepLearning_TabularDataAugmentation_Refactored.html
The variables in that case are continuous. Or did I missunderstand understand your question?
dta supports continuous data as input variables: https://lschmiddey.github.io/fastpages_/2021/11/07/DeepLearning_TabularDataAugmentation_Refactored.html The variables in that case are continuous. Or did I missunderstand understand your question?
Thanks for your reply!
Yes, the variables are continious, but the target is binary. In that case we need to specify target_name and target_class for Loader class. So i want to give continious or binary class (which leads to result in my attachement). My question is how i can specify target_class with continious variable.
Good question, I did build the dta for binary/categorical variables and havent thought about this. I will have a look whether I can update dta to also include continuous target variables
Ahoi Dimot322,
I updated the package, so one can now also use it in cases when there is no target per se required/wanted. You can install the package like so: pip install deep-tabular-augmentation
The latest version under which this feature is available is 0.5.0
Also I attached a notebook file where you can see this in action: https://github.com/lschmiddey/deep_tabular_augmentation/blob/main/Notebooks/DeepLearning_DataAugmentation_RF_2.ipynb
I hope this helps :)
Thanks, it helps a lot!
Ahoi Dimot322, I updated the package, so one can now also use it in cases when there is no target per se required/wanted. You can install the package like so: pip install deep-tabular-augmentation The latest version under which this feature is available is 0.5.0 Also I attached a notebook file where you can see this in action: https://github.com/lschmiddey/deep_tabular_augmentation/blob/main/Notebooks/DeepLearning_DataAugmentation_RF_2.ipynb I hope this helps :)