Hello,
I'm very interested in your project, but it seems a bit unclear to me. Which of these statements is true?
1) You want to run Deep learning algorithm by yourself and need a consultation
2) You want the deep learning algo to be written for you and the result - the excel sheet to be provided
3) You want to fill the empty cells with data and algo doesn't really matter (because there are more other algos for this purpose)
To be more precise, what is more important for you: the result or the algo?
I have inspected your data:
1) "Farbwerke" column has only 7 values missed out of 1620 so it can be recovered
2) "Papierformat (Breite) (Max.)" has 220 missed values out of 1620, it's about 14% and still can be recovered
3) "Papierformat (Höhe) (Max.)" has as 221 missed values out of 1620, it's about 14% and still can be recovered
4) "Druckformat (Breite) (Max.)" has 794 out of 1620, it's about 49% and normally can't be recovered
5) "Druckformat (Höhe) (Max.)" has 790 out of 1620 it's about 49% and normally can't be recovered
Best regards, Sofya