Hello, Pablo.
I have a question. What is the difference between overfitting and not overfitting?
If I choose genes by the BDTNP data and geometry, I am overfitting the insitu data.
If I choose genes by DGE data, I am not overfitting. Am I correct?
thanks
Kai He
Created by Kai He hekind Exactly you cant, you have to choose the genes initially wo parameter search Hi,
Thanks for your replay. I am confused with one thing. I can't search parameters of model by all 84 genes. I can only use 20/40/60 genes and all DGE data.
I just want to make sure if I violate the rules of competition. Hi I am sorry but overfitting does not have to do with what you said.
The point is having the solution and having to predict the solution, so you can use DGE to define the position of each cell, but your result is not generalizable.
P