Deep Learning
Autoencoders are one of the most popular supervised learning methods.
False
Proposed by @svpino
Card 1 of 276. Answer: False
False
Proposed by @svpino
Card 1 of 276. Answer: False
About · Quizzes · History of AI · To suggest new questions or report an error send me a dm.
False. Autoencoders are unsupervised, or more precisely self-supervised: they learn to reconstruct their own input, so the target is the input itself and no human labels are needed. Supervised learning needs input-label pairs (x, y), such as photos tagged with what they show.
The loss shows the difference. A supervised classifier minimizes L(f(x), y); an autoencoder minimizes L(g(f(x)), x), with f the encoder and g the decoder. That made autoencoders a pretraining tool: Hinton and Salakhutdinov's 2006 Science paper trained deep autoencoders on unlabeled data, and their codes beat linear methods such as PCA at compressing images and documents. A classifier can then be trained on the codes with few labels, which is semi-supervised learning.
A common follow-up: a linear autoencoder trained with squared error learns the same subspace as the top principal components, as Baldi and Hornik showed in 1989, so non-linear activations are what let it do more. Masked autoencoders, which hide 75% of an image's patches and rebuild them, are a modern self-supervised pretraining method.