A deep belief network (DBN) is a class of deep neural network, composed of multiple layers of hidden units, with connections between the layers; where a DBN differs is these hidden units don't interact with other units within each layer. A DBN can learn to probabilistically reconstruct its input without supervision, when trained, using a set of training datasets. It is a joint (multivariate) distributions over large numbers of random variables that interact with each other. These representations sit at the intersection of statistics and computer science, relying on concepts from probability theory, graph algorithms, machine learning, and more.
The layers act as feature detectors. After the training step, a DBN can be trained with supervision to perform classification.
We will be covering the following chapters in the chapter:
- Understanding deep belief networks
- Model training
- Predicting the label
- Finding the accuracy of the model
- DBN implementation for the MNIST...