Currently, we are choosing hyperparameters in a sequential manner. An alternative would be a fully systematic search in style of "Optimized parameter selection reveals trends in Markov state models for protein folding" (Husic et al, 2016).
We should briefly discuss both approaches and clearly describe how we are choosing our key hyperparameters.
Currently, we are choosing hyperparameters in a sequential manner. An alternative would be a fully systematic search in style of "Optimized parameter selection reveals trends in Markov state models for protein folding" (Husic et al, 2016).
We should briefly discuss both approaches and clearly describe how we are choosing our key hyperparameters.