# Function index

MatConvNet includes several MATLAB functions organized as follows:

• Building blocks. These functions implement the CNN computational blocks that can be combined either manually or using one of the provided wrappers to construct CNNs.
• SimpleCNN wrapper. SimpleNN is a lightweight wrapper implementing CNNs that are linear chains of computational blocks.
• DagNN wrapper. DagNN is an object-oriented wrapper supporting more complex network topologies.
• Other functions. These helper functions are used to initialize and compile MatConvNet.

There is no general training function as training depends on the dataset and problem. Look at the examples subdirectory for code showing how to train CNNs.