I have a dataset (dataframe) with 15 input variables and 10 output variables which can be either categorical or numeric variables. ( I .I treated in learn how to do that in both cases using tensorflow or keras). I want to develop a deep learning model that can, based on a vector (1,15), be able to obtain an output one (1,10). Input=[[1,12,3,0.3.....]] numerical variables. Output =[[0.3,0.2,1,3....]] As a numerical output or [[1,2,1,0..]] as a categorical variable
I have reviewed and an lstm or cnn model could be used for my case, but when structuring it it has been somewhat confusing to develop a model that generates an output vector of that length, as in typical cases where for n inputs only the prediction o is performed. classification for an attribute or output variable. Some have an idea of how to develop such a structure that can have multiple outputs