Autoencoder – A type of unsupervised neural network that learns efficient representations of input data by encoding and then decoding it. Autoencoders are widely used for anomaly detection, data compression, and noise reduction in images and signals.