Start free trial
Take Naologic for a spin today, no credit card needed and no obligations.
Start free trial

Autoencoder - When should we use autoencoders?


When labels are not available, autoencoders can be used to find additional classes. Training a model to effectively partition data into sub-classes can be achieved by utilizing a small subset of the data, say, with only two or three output neurons. This is very useful for clustering and unsupervised learning.