Skip to content
#

mixed-loss-variational-autoencoder

Here is 1 public repository matching this topic...

Training VAE and an Mixed-Loss VAE with normal network traffic, then implementing validation and test steps that fold in the attack traffic to ensure models' reconstruction error satisfactorily separates data at with accuracy. Leveraging Bayesian optimization to tune hyperparameters and testing one anomaly at a time hold out strategy with AUCs.

  • Updated Jun 14, 2024
  • Jupyter Notebook

Add this topic to your repo

To associate your repository with the mixed-loss-variational-autoencoder topic, visit your repo's landing page and select "manage topics."

Learn more