New answers tagged hyperparameter
0
Unfortunately I don't think there is an easy way of doing this. It all comes down to this: you need some way to evaluate how good your GAN is, which isn't an easy task.
Why not use the discriminator?
The way evaluation is done during training is through the discriminator. So, one first thought would be to hold out one part of the training samples, generate ...
7
How to decide whether to optimize model hyperparameters on a development set or by cross-validation?
The development set, in other words, validation set or holdout set is like executing cross-validation for one fold only. It's faster, therefore cheaper. Typically, cross-validation is more robust because the performance is averaged across several tests instead of one. This might make sense when dealing with smaller datasets. When, data is not a problem, ...
Top 50 recent answers are included
Related Tags
hyperparameter × 498machine-learning × 195
cross-validation × 126
optimization × 99
neural-networks × 94
svm × 39
scikit-learn × 36
classification × 32
bayesian × 27
model-selection × 27
gaussian-process × 27
python × 23
bayesian-optimization × 23
random-forest × 22
regression × 21
conv-neural-network × 16
feature-selection × 15
validation × 15
overfitting × 15
tuning × 15
xgboost × 14
regularization × 11
model-evaluation × 11
r × 10
time-series × 10