This program - Steel Knowledge-Yielding Network - is a graphical user interface intended for easy to use of arbitrary Bayesian artificial neural networks within MAP library. Using this program, the following three items can be checked.
Significane for each input of each model can be visualised.
The changes of the result value, uncertainty, and noise as a function of input variables.
An optimized input value that derives a result close to the target value using the genetic algorithm.
Each model included in this program was built using Bigback5 code created by David Mackay.
This software is graphical user interface for easy use of arbitrary artificial neural network models in the MAP labrary. Each model is the output via Neuromat Model Manager by Thomas Sourmail, and the artificial neural network is built using the code Bigback5 by David Mackay. Each model includes c/_c*, c/_w*, c/_w*.lu, committee, labels, max_noise, MINMAX, and read me files.
README
A text file containing the details of the model.
MINMAX
A text file containing the minimum and maximum limits of each input and output variable. This file is used to normalise and unnormalise the input and output data.
labels
A text file containing the labels of input variables.
committee
A text file containing the list of committee for model.
max_noise
A text file containing the miximum of noise. If there is no max_noise file, then the noise value is evaluated using sigma_nu value in _c* files.
_w*f
The weights files for the different models.
*.lu
Files containing information for calculating the size of the error bars for the different models.
_c*
Files containing information about the perceived significance value for each model.
_R*
Files containing values for the noise, test error and log predictive error for each model.
Watch the following movie to assist in the use the program.
![[A logo showing the University of Cambridge Crest]](https://www.phase-trans.msm.cam.ac.uk/Univ.jpg)
