On-Line Demo ::

PNN Web Client

Regression type:
Max Degree:
Max Number of Terms:
Robust Calculations:
Number of Validation Rows:
Initial settings
Row value delimiter      


:: User's information

  You can try Polynomial neural network generic algorithm with examples mentioned in the article.
Load example data:
      - Random noise with three more random variables added
      - Prediction of the sublimation enthalpy
Then click Start button.
  This page helps PNN users initialize PNN generic algorithm using buffer input. Use copy and paste method to place the initial array into text area above. Select the appropriate row value delimiter from the drop-down list below text area.
Important. Each row should start with new line. The last value in each row should contain the target value. Use simbol "." as decimal point separator.
  Use Configuration section above text area to change the settings (parameters) for PNN generic algorithm
  Check the Robust checkbox to enable the robust iteration during models calculation
  Press Start button to start the calculation and view the results
  Press Clear button to return to default algorithm settings and zero-field table of minimum size
  Limitation. Although the commercial version of PNN engine has no limitations on number of input rows and columns, this demo version has limit of maximum 10 attributes and 15 rows.
  To learn more about Polynomial neural network generic algorithm, please read our article
  Please, don't hesitate to contact us about additional Polynomial neural network generic algorithm information.


Copyright © 2002-2009, PNN Ltd. All rights reserved.