Anomaly Detection by Reconstruction Error
Flag outliers the model cannot rebuild.
Spotting the Odd One Out
Autoencoders are great at finding anomalies: data points that do not fit the normal pattern. The idea is clever and surprisingly simple.
Train on Normal Only
You train the autoencoder on normal data alone. It becomes an expert at rebuilding the kinds of inputs it has seen many times.
All lessons in this course
- Encoder, Bottleneck & Decoder
- Denoising Autoencoders
- Variational Autoencoders & the Latent Space
- Anomaly Detection by Reconstruction Error