Two methods for studying the consistency problems of a class of binary relation inference networks are described. One method is derived using the mathematical concepts of energy function (Et) and delta energy function (ΔEt), where both functions have closely related geometrical interpretations. By properly formulating ΔEt as matrix quadratic form, network convergence is shown to be directly related to the matrix property of negative semidefiniteness. The other method, which can be applied in either a discrete-time or continuous-time framework, is based on studying the eigenvalue problem for an associated state-space model of the inference network. The merits and limitations of the proposed methods are discussed, with reference to several specific examples.
References
-
-
1)
-
S.Y. Kung ,
K. Hwang ,
D. DeGroot
.
(1989)
VLSI array processors for signal/image processing, Parallel processing for supercomputers and artificial intelligence.
-
2)
-
Lam, K.P. and Su, C.: ‘Theory and applications ofbinary relation inference network’.Submitted manuscript, 1995.
-
3)
-
Goddard, N.H., Lynne, K.J., Mintz, T.: `Rochester connectionist simulator (RCS)', 233 (revised), Technical Report, 1988.
-
4)
-
A.J. Laub ,
J.N. Little
.
(1986)
Control system toolbox user's guide: for use with MATLAB.
-
5)
-
K.P. Lam ,
A.H. Levis ,
H.E. Stephanou
.
(1992)
Energy functions and convergence analysis of inference networks, Distributed intelligence systems.
-
6)
-
G. Strang
.
(1988)
Linear algebra and its applications,,,.
-
7)
-
J.J. Hopfield
.
Neurons with graded response have collective computational propertieslikethose of two-state neurons.
Proc. Natl. Acad. Sci., USA
,
3088 -
3092
-
8)
-
A.K. Mackworth ,
S. Shapiro
.
(1987)
Constraint satisfaction, Encyclopedia of artificial intelligence.
-
9)
-
Su, C.: `A binary relation inference network for constrained optimisation', 1992, PhD, University of British Columbia.
-
10)
-
A.N. Michel ,
J.A. Farrell ,
W. Porod
.
Qualitative analysis of neural networks.
IEEE Trans.
,
229 -
243
-
11)
-
Lam, K.P.: `Using hybrid techniques for a continuous-time inference network', Proceedings of the IEEE international symposium on Circuitsand systems, 1991, Singapore, p. 1721–1724.
http://iet.metastore.ingenta.com/content/journals/10.1049/ip-cta_19960438
Related content
content/journals/10.1049/ip-cta_19960438
pub_keyword,iet_inspecKeyword,pub_concept
6
6