Contents

Autoencoders

An autoencoder is a special type of Feedforward Networks which

  • encodes its nput x_i into a hidden representation h
  • Decodes the input again from this hidden representation

Types of Autoencoders

Based on completeness

Under complete

An auto encoder where dim(h) < dim(x_i) is called an under complete autoencoder

Over complete

An auto encoder where dim(h) > dim(x_i) is called an under complete autoencoder Why required? Special treatment required to remove the possibility of duplication?

Function