Top Page | Upper Page | Contents | About This Site | JAPANESE

Auto Encoder

Auto Encoder is one of the Neural Network . It defines the intermediate layers that input data and output data are same.

It is used as the tool of Deep Learning to delete noise and to find characteristics of the data.

Boltzmann Machine

Boltzmann Machine is one of the graphical modeling. It is made in the field of Information Statistical Mechanics.

It uses the idea of energy as the parameter of optimisation.

Restricted Boltzmann Machine : RBM

In original Boltzman Machine, there is not the kinds of variables. RBM divides them into inputs and outputs. And we put the rule that there are no link among same group.

RBM has same function of Auto Encoder.

Analysis Using Intermediate Layer

NEXT Convolutional Neural Network (CNN)