Neural Network Toolbox™ provides algorithms, pretrained models, and apps to create, train, visualize, and simulate both shallow and deep neural networks. You can perform classification, regression, clustering, dimensionality reduction, time-series forecasting, and dynamic system modeling and control.
Deep learning networks include convolutional neural networks (ConvNets, CNNs) and autoencoders for image classification, regression, and feature learning.
For small training sets, you can quickly apply deep learning by performing transfer learning with pretrained deep networks. To speed up training on large datasets, you can use Parallel Computing Toolbox™ to distribute computations and data across multicore processors and GPUs on the desktop, and you can scale up to clusters and clouds (including Amazon EC2® P2 GPU instances) with MATLAB® Distributed Computing Server™.
Learn the basics of Neural Network Toolbox
Construct and train convolutional neural networks (CNNs, ConvNets) for classification and regression and autoencoder neural networks for learning features
Create a neural network to generalize nonlinear relationships between example inputs and outputs
Train a neural network to generalize from example inputs and their classes, construct a deep network using autoencoders
Discover natural distributions, categories, and category relationships
Model nonlinear dynamic systems; make predictions using sequential data
Control nonlinear systems using model-predictive, NARMA-L2, and model-reference neural networks
Define new neural network architectures and algorithms for advanced applications