Autoencoders

Question 1

What is a high level definition of an autoencoder?

Answer 1

an unsupervised learning technique in which we leverage neural networks for the task of representation learning

Question 2

at a high level how does an autoencoder learn the representation?

Answer 2

the neural network has a bottleneck inside of it that forces a compressed knowledge representation of the original input.

Question 3

at a medium level how does an autoencoder work?

Answer 3

we take an unlabeled dataset and frame it as a supervised learning problem, tasked with outputting a reconstruction of the original input via a bottlenecked network. We train our network, minimizing the reconstruction error