Network Recall

Neural network recall is the processing of new inputs through a trained network. Qnet offers a variety of methods for accessing Qnet trained networks. These include the recall mode available within Qnet, C source and DLL access for programmers, and QnetTool for integration of trained networks into Windows based spreadsheet and data analysis applications.

To access your neural network from within Qnet, simply open your network definition file and supply the file containing the input node data in the recall setup dialog. Use the same file format rules that are used for training data. If targets are provided in recall mode (optional), the predictive qualities of a model can be checked. After all input patterns are processed, the network’s output may be analyzed using NetGraph and Qnet’s information viewer. Network outputs can also be saved to an ASCII file with tab delimited data columns for importing into spreadsheet applications for further analysis.