DANA [Design Advisor/Neural Analyzer] is a PC-based experimental design advisor and neural network analysis system for understanding and optimizing physical processes. Given a specified number of process inputs and outputs, the Design Advisor gives two options for a set of experiments to perform in order to map the input variables to the process output responses: a standard statistical design and a novel smaller design specifically developed for neural network analysis. The neural network employed in the Neural Analyzer is the back-propagation type, in which a single hidden layer is used to relate the input nodes (variables) to the output nodes (responses) for the experimental data set analyzed. The Network Output module uses the neural model to produce a number of useful graphs and a "Virtual Process" for interactive modeling.
[1]
Geoffrey E. Hinton,et al.
Learning internal representations by error propagation
,
1986
.
[2]
Sidney Addelman,et al.
trans-Dimethanolbis(1,1,1-trifluoro-5,5-dimethylhexane-2,4-dionato)zinc(II)
,
2008,
Acta crystallographica. Section E, Structure reports online.
[3]
Anders Krogh,et al.
Introduction to the theory of neural computation
,
1994,
The advanced book program.
[4]
W. Press,et al.
Numerical Recipes: The Art of Scientific Computing
,
1987
.
[5]
George Cybenko,et al.
Approximation by superpositions of a sigmoidal function
,
1992,
Math. Control. Signals Syst..