Patent
1993-06-17
1995-10-24
Downs, Robert W.
395 21, 395 23, G06F 1580, G06K 962
Patent
active
054616980
ABSTRACT:
Given a set of objects (A, B, C, . . . ), each described by a set of attribute values, and given a classification of these objects into categories, a similarity function accounts well for this classification when only a small number of objects are not correctly classified. A method for modelling a similarity function using a neural network comprises the steps of: (a) inputting feature vectors to a raw input stage of a neural network respectively for object S in the given category, for other objects G in the same category being compared the object S, and for object B outside the given category; (b) coupling the raw inputs of feature vectors for S, G, and B to an input layer of the neural network performing respective set operations required for the similarity function so as to have a property of monotonicity; (c) coupling the input elements of the input layer to respective processing elements of an hidden layer of the neural network for computing similarity function results adaptively with different values of a coefficient w of the similarity function; (d) coupling the processing elements of the hidden layer to respective output elements of an output layer of the neural network for providing respective outputs of an error function measuring the extent to which object S is more similar to object G than to object B; and (e) obtaining an optimal coefficient w by back propagation through the neural network which minimizes the error outputs of the error function.
REFERENCES:
patent: 4319221 (1982-03-01), Sakoe
patent: 4326101 (1982-04-01), Sakoe
patent: 4467437 (1984-08-01), Tsuruta et al.
patent: 4484348 (1984-11-01), Shizuno
patent: 4528688 (1985-07-01), Ichikawa et al.
patent: 4530110 (1985-07-01), Nojiri et al.
patent: 4559604 (1985-12-01), Ichikawa et al.
patent: 4624011 (1986-11-01), Watanabe et al.
patent: 4651289 (1987-03-01), Maeda et al.
patent: 4716593 (1987-12-01), Hirai et al.
patent: 4723290 (1988-02-01), Watanabe et al.
patent: 4731845 (1988-03-01), Matsuki et al.
patent: 4745562 (1988-05-01), Prazdny
patent: 4752957 (1988-06-01), Maeda
patent: 4773099 (1988-09-01), Bokser
patent: 4783806 (1988-11-01), Nakamura et al.
patent: 4876731 (1989-10-01), Loris et al.
patent: 4893255 (1990-01-01), Tomlinson, Jr.
patent: 4897811 (1990-01-01), Scofield
patent: 5109431 (1992-04-01), Nishiya et al.
Lippmann, R. P., "An Introduction to Computing With Neural Nets", IEEE ASSP Mag., Apr. 1987, 4-22.
"Pattern Classification Using Trainable Logic Networks", Evans, IJCNN Int'l Joint Conference on Neural Networks, 19-22 Jun. 1989, Washington, US, pp. 429-433.
"Neuronale Netze als Signal--und Musterklassifikatoren", Barschdorff et al., Technisches Messen TM, vol. 57, No. 11, Nov. 1990, Munich, Germany, pp. 437-444.
"Empirically Guided Software Development Using Metric-Based Classification Trees", Porter et al., IEEE Software, Mar. 1990, pp. 46-54.
"Change Management for Very Large Software Systems", Maarek et al., Seventh Ann. Int'l Phoenix Conf. on Computers & Comm., Scottsdale, Ariz., Mar. 1988, pp. 280-285.
"Information Distribution Aspects of Design Methodology", Parnas, Information Processing 71, North-Holland Publishing Company (1972), pp. 339-344.
Paper entitled "On the Criteria To Be Used In Decomposing Systems Into Modules", by Parnas, (Aug. 1971).
Ada 9X Program Structure and Compilation Issues, Draft Mapping Document, 21 Feb. 1991, pp. 10-1--10-5.
"Software Change Dynamics or Half of all Ada Compilations are Redundant", Adams et al., Univ. of Karlsruhe, Karlsruhe, FRG, Jun. 8, 1989, pp. 1-15.
"System Structure Analysis: Clustering with Data Bindings", Hutchens et al., IEEE Transactions on Software Eng., vol. SE-11, No. 8, Aug. 1985, pp. 749-757.
"System Partitioning and Its Measure", Belady et al., The Journal of Systems & Software 2, 23-29 (1981), Elsevier North Holland, Inc., 1981, pp. 23-29.
"Features of Similarity", Tversky, Psychological Review, vol. 84, No. 4, Jul. 1977, pp. 327-351.
"A Parallel Network that Learns to Play Backgammon", Tesauro et al., Artificial Intelligence 39 (1989), pp. 357-390.
"Modularization and Hierarchy in a Family of Operating Systems", Habermann et al., Communications of the ACM, May 1976, vol. 19, No. 5, pp. 266-272.
"Version Management in Gypsy", Cohen et al., Proceedings of the ACM SIGSOFT/SIGPLAN . . . , Boston, Mass., Nov. 28-30, 1988, pp. 201-215.
"Error Localization During Software Maintenance: Generating Hierarchical System Descriptions from the Source Code Alone", Selby et al., Conference on Software Maintenance-1988, Phoenix, Ariz., Oct. 24-27, 1988, pp. 192-197.
"Extracting and Restructuring the Design of Large Systems", Choi et al., IEEE Software, Jan. 1990, pp. 66-71.
Article entitled "Tools for Software Configuration Management" by Tichy, Proc. Int'l Workshop on Software Conversion and Configuration Control (Jan. 1988), pp. 1-20.
"Configuration Management, The Notion and the Tools", Estublier, Proceedings of Jan. 27-29, 1988 Grassau, pp. 38-61.
"Program-Variations-in-the-Small", Winkler et al., Proceedings of the International Workshop on Software Version and Configuration Control, Jan. 27-29, 1988 Grassau, pp. 175-196.
"Learning Internal Representations by Error Propagation", Rumelhart, pp. 318-362; MIT Press, (1986).
"A Taxonomy of Some Principal Types of Data and of Multidimensional Methods for Their Analysis", Shepard, pp. 21-47, (1970).
"Programming-in-the Large Versus Programming-in-the-Small", DeRemer et al., IEEE Transactions on Software Engineering, vol. SE-2, No. 2, Jun. 1976, pp. 80-86.
"The C Information Abstraction System", Chen et al., IEEE Transactions on Software Engineering, vol. 16, No. 3, Mar. 1990, pp. 325-334.
"Configuration Management in BiiN.TM. SMS", Schwanke et al., ACM, (1989), pp. 383-393.
"A Method for Improving the Classification Speed of Clustering Algorithms Which Use a Euclidean Distance Metric", Curle et al., Proceedings of the IEEE, vol. 69, No. 1, Jan. 1981, pp. 128-129.
Hanson Stephen J.
Schwanke Robert W.
Ahmed Adel A.
Downs Robert W.
Siemens Corporate Research Inc.
LandOfFree
Method for modelling similarity function using neural network does not yet have a rating. At this time, there are no reviews or comments for this patent.
If you have personal experience with Method for modelling similarity function using neural network, we encourage you to share that experience with our LandOfFree.com community. Your opinion is very important and Method for modelling similarity function using neural network will most certainly appreciate the feedback.
Profile ID: LFUS-PAI-O-1892765