Developing Neural Network Applications

This three-day neural computing course focuses on the practical aspects of creating and deploying neural network applications using NeuralWare development tools. It provides an overview of the NeuralWare product line, and describes the software architecture of each application and library. It also identifies the functional and architecture relationship among the products. Next, the course presents an overview of the methods that can be used to deploy neural network models, including standard NeuralWare .nnd and .npr files, FlashCode(tm), Designer Pack, and use of NeuralWare library components.

During the course, several example models will be built using NeuralWare products, and demonstration applications will be developed using C and Microsoft Visual Basic.

Syllabus

Day 1 Products and Architectures

- NeuralWorks
      Professional II/PLUS
      Predict
      Designer Pack
      User-Defined Neuro-Dynamics
- Predict Run-Time Kit (RTK)
- Predict Software Development Kit (SDK)
- NeuralStudio
      Library Components
      Java Components
Lab Exercises - Building and Deploying Models using NeuralWorks Products


Day 2 Advanced Model Deployment Options

- Trained Models
      Excel Spreadsheets
      FlashCode
      Visual Basic
- Models capable of continuous learning
      Designer Pack
      SDK
      Components
Lab Exercise - Building Models with Command Line Interface


Day 3 Special Topics

- Windows Programming Issues
- UNIX/Linux Programming Issues
- Self-Organizing Map Models
      Clusters
      Visualization
- Project Data Analysis and Project Management