usenet@cps3xx.UUCP (Usenet file owner) (10/14/89)
I'm looking for papers about subgraph isomorphism/
best matching subgraph using neural nets. So far I
have 1 reference:
Paul Cooper, "Structure Recognition by Connectionist Relaxation:
Formal Analysis", Proc. CSCSI-88.
Any other references would be greatly appreciated; if
there are enough responses I will summarize. Please
email me at: raja%frith.egr.msu.edu@uunet.uu.net
OR raja%frith.egr.msu.edu@cunyvm.cuny.edu
Thank you.
Narayan Sriranga Raja.
________ ________ __ __ ________
| | | | (/ \/ ) | |
| | | | / | |
| | | | / | |
| | | | / | |
| | | | _/ __ | | (Raja)
| | | | (__/ \) | |
/usenet@cps3xx.UUCP (Usenet file owner) (10/26/89)
Here's the summary of responses I received.
Raja.
-------------------------------------------------------------
From zador-anthony@YALE.ARPA Fri Oct 13 18:35:23 1989
Gene Gindi, Eric Mjolsness, P. Anandan, Joachim Utans, and Tony Zador
have in various combinations presented papers on vision using
graph-matching neural nets.
The best ref is a paper by the 1st three of these, which appeared in
Neural Computation (T. Sejnowski's journal, i think this is what it's
called), Summer, 1989
--------------------------------------------------------------
From: jagota@cs.Buffalo.EDU (Arun Jagota)
[PS - I do have a reference to applying connectionist models
to some other graph problems. This work was done by two MS
students under Michael Luby at U. Toronto.]
On the Computational Complexity of Finding Stable State Vectors in
Connectionist Models (Hopfield Nets).
Gail H. Godbeer, John Lipscomb, Michael Luby
TR 208/88 March, 88 U. Toronto
]
---------------------------------------------------------------
From Todd.Kaufmann@NL.CS.CMU.EDU Mon Oct 16 09:48:32 1989
In the past couple months in Science there was a short paper on a NN
that could do graph planarization.. perhaps some of the references
there would lead you to other graph algorithms.
-----------------------------------------------------------------
From: platt@synaptics.com (John Platt)
The folks at Yale have been looking at NNs used for graph
problems. I think you'd be interested in the report
"Stickville: A Neural Net for Object Recognition via Graph Matching"
by Grant Shumaker, Gene Gindi, Eric Mjolsness, and P. Anandan,
Report 8908, Center for Systems Science, Electrical Engineering,
Yale University, April 1989.
------------------------------------------------------------------
Sender: mackeown%compsci.bristol.ac.uk@NSFnet-Relay.AC.UK
Here's another reference on graph-matching which may interest you:
Von der Malsburg, C. "Pattern Recognition by Labelled Graph
Matching", Neural Networks, Vol.1, No.2, pp.141-148, 1988.
------------------------------------------------------------------
________ ________ __ __ ________
| | | | (/ \/ ) | |
| | | | / | |
| | | | / | |
| | | | / | |
| | | | _/ __ | | (Raja)
| | | | (__/ \) | |
/