Learn to use the state-of-the-art neural net framework in Wolfram Language. Find pre-built and pre-trained models in the Wolfram Neural Net Repository and see how to adapt them to your own applications.
This course provides an introduction to the state-of-the-art Neural Net Framework in Wolfram Language. You will learn how to explore the Wolfram Neural Net Repository for pre-built and pre-trained models and how to apply them to your own dataset. Also see how you can use transfer learning to adapt models for your own applications. The course will then discuss the building blocks of neural nets, along with instructions for putting them together within a symbolic framework to build your own neural network. Simple examples of training and testing a neural network will be discussed, along with options to inspect output from hidden layers of the net.
Featured Products & Technologies: Wolfram Language (available in Mathematica and Wolfram|One)
Outline
What Is a Neural Network: Get introduced to the concepts of neural net architecture and deep learning, as well as the high-level overview of what neural nets are doing.
Building Blocks of the Neural Net Framework: Learn about the building blocks of neural networks available within the Wolfram Neural Net Framework. Learn about how data is represented, encoders and decoders, built-in layers and the tools available for training and testing networks.
Explore Examples: Explore image, audio, video and natural language processing. Practical examples include a neural network to recognize handwritten digits from their images.
The Neural Net Repository: Take a quick tour of the repository, a public collection of neural networks, from which models can be retrieved and reused with little effort. See a simple example of using transfer learning to adapt and retrain the pre-built Wolfram ImageIdentify net model for the specific use case of determining dog breed from an image.
This course provides an introduction to the state-of-the-art Neural Net Framework in Wolfram Language. You will learn how to explore the Wolfram Neural Net Repository for pre-built and pre-trained models and how to apply them to your own dataset. Also see how you can use transfer learning to adapt models for your own applications. The course will then discuss the building blocks of neural nets, along with instructions for putting them together within a symbolic framework to build your own neural network. Simple examples of training and testing a neural network will be discussed, along with options to inspect output from hidden layers of the net.
Featured Products & Technologies: Wolfram Language (available in Mathematica and Wolfram|One)
Outline
What Is a Neural Network: Get introduced to the concepts of neural net architecture and deep learning, as well as the high-level overview of what neural nets are doing.
Building Blocks of the Neural Net Framework: Learn about the building blocks of neural networks available within the Wolfram Neural Net Framework. Learn about how data is represented, encoders and decoders, built-in layers and the tools available for training and testing networks.
Explore Examples: Explore image, audio, video and natural language processing. Practical examples include a neural network to recognize handwritten digits from their images.
The Neural Net Repository: Take a quick tour of the repository, a public collection of neural networks, from which models can be retrieved and reused with little effort. See a simple example of using transfer learning to adapt and retrain the pre-built Wolfram ImageIdentify net model for the specific use case of determining dog breed from an image.
