search

theislab / cpa

The Compositional Perturbation Autoencoder (CPA) is a deep generative framework to learn effects of perturbations at the single-cell level. CPA performs OOD predictions of unseen combinations of drugs, learns interpretable embeddings, estimates dose-response curves, and provides uncertainty estimates.
BSD 3-Clause "New" or "Revised" License
76 stars 17 forks source link

Hyperparameter optimization #10

Closed yanwu2014 closed 3 months ago

yanwu2014 commented 1 year ago

The provided tutorials seem to have very specific hyperparameters tuned to the dataset. Since we're interested in running CPA on new data, do you have recommendations for which hyperparemeters are key for CPA performance?

tuln128 commented 1 year ago

Not sure if this site is still being supported. However, I have a similar question related to hyperparameter tuning for the CPA model, which has been posted here: https://github.com/theislab/cpa/issues/7#issuecomment-1260240287 Hope that someone can help make this matter clear. Thanks in advance,

ArianAmani commented 3 months ago

Hyperparameter tuner added #46 https://github.com/theislab/cpa/#how-to-optmize-cpa-hyperparamters-for-your-data