[R] Autoencoder or Variational Autoencoder for finding anomalies in a dataset in comparison with a real-world dataset

I have a general question about Autoencoders (AE) or Variational Autoencoders (VAE). I possess both a real-world dataset and a synthetic dataset, and my goal is to identify discrepancies in the synthetic dataset compared to the real-world dataset. While existing research focuses on anomaly detection within a dataset using AEs, I am specifically interested in detecting anomalies in the synthetic dataset when compared to the real-world dataset. I am wondering if there are any papers addressing this issue. Additionally, I …

autoencoder autoencoders comparison dataset general identify machinelearning question research synthetic vae variational autoencoders world