Combining linear discriminant functions with neural networks for supervised learning

@article{Chen1997CombiningLD,
  title={Combining linear discriminant functions with neural networks for supervised learning},
  author={Ke Chen and Xiang Yu and Huisheng Chi},
  journal={Neural Computing & Applications},
  year={1997},
  volume={6},
  pages={19-41}
}
A novel supervised learning method is proposed by combining linear discriminant functions with neural networks. The proposed method results in a tree-structured hybrid architecture. Due to constructive learning, the binary tree hierarchical architecture is automatically generated by a controlled growing process for a specific supervised learning task. Unlike the classic decision tree, the linear discriminant functions are merely employed in the intermediate level of the tree for heuristically… CONTINUE READING

Citations

Publications citing this paper.
Showing 1-10 of 13 extracted citations

Multiresolution Learning on Neural Network Classifiers: A Systematic Approach

2009 International Conference on Network-Based Information Systems • 2009
View 3 Excerpts
Highly Influenced

Deep and Modular Neural Networks

Handbook of Computational Intelligence • 2015

References

Publications referenced by this paper.
Showing 1-10 of 64 references

Dynamic error propagation networks

A. J. Robinson
Ph.D. Thesis, University of Cambridge, • 1989
View 8 Excerpts
Highly Influenced

Learning to tell two spirals apart

K. J. Lang, M. J. Witbrock
D. Touretzky, G. Hinton, and T. Sejnowski, editors, Proc. The 1988 Connectionist Models Summer School, pages 52{59, San Mateo, CA, • 1989
View 4 Excerpts
Highly Influenced

Pattern classification and scene analysis

A Wiley-Interscience publication • 1973
View 6 Excerpts
Highly Influenced

On growing better decision trees from data

K.V.S. Murthy
Ph.D. Thesis, The Johns Hopkins University, • 1995
View 4 Excerpts
Highly Influenced

Hierarchical Mixtures of Experts and the EM Algorithm

Neural Computation • 1994
View 16 Excerpts
Highly Influenced

Learning in networks is hard

S. Judd
Proc. IEEE Int. Conf. Neural Networks, volume 2, pages 685{692, • 1987
View 4 Excerpts
Highly Influenced

A modi ed HME architecture for text-dependent speaker identi cation

K. Chen, D. H. Xie, H. S. Chi
IEEE Transactions on Neural Networks, 7(5):1309{1313, • 1996

Similar Papers

Loading similar papers…