Autoencoders

Neural networks that learn to compress input data into a compact representation and reconstruct it, useful for dimensionality reduction, denoising, and anomaly detection.