Abstract : This research is to develop a self-improving system that is capable of performing concept formation for associating materials and process design using inductive coupling techniques. The first phase of the work focuses on developing an artificial neural network learning for function approximation. The objective for neural network function approximation is to learn the input-output mapping efficiently. The second phase of the work focuses on developing an artificial neural network learning algorithm for time-series prediction. The third phase of the work focuses on model selection. We have successfully used both neural networks approach and neuro-fuzzy approach to select the most important features (attributes) for function mapping and construct a model with a minimum generalization error. We have applied this approach, successfully, identifying material properties and predicting ternary systems compounding formation.
[1]
C. L. Philip Chen,et al.
Materials structure-property prediction using a self-architecting neural network
,
1998
.
[2]
C. L. Philip Chen,et al.
A rapid supervised learning neural network for function interpolation and approximation
,
1996,
IEEE Trans. Neural Networks.
[3]
C. L. Philip Chen,et al.
Feature sequencing in the rapid design system using a genetic algorithm
,
1996,
J. Intell. Manuf..
[4]
C. L. Philip Chen,et al.
A rapid learning and dynamic stepwise updating algorithm for flat neural networks and the application to time-series prediction
,
1999,
IEEE Trans. Syst. Man Cybern. Part B.
[5]
C. L. Philip Chen,et al.
Integration of design and manufacturing: solving setup generation and feature sequencing using an unsupervised-learning approach
,
1994,
Comput. Aided Des..