{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,11,24]],"date-time":"2025-11-24T07:08:14Z","timestamp":1763968094381},"reference-count":24,"publisher":"MIT Press - Journals","issue":"4","license":[{"start":{"date-parts":[[2020,7,15]],"date-time":"2020-07-15T00:00:00Z","timestamp":1594771200000},"content-version":"vor","delay-in-days":2662,"URL":"https:\/\/creativecommons.org\/licenses\/by-nc\/4.0\/"}],"content-domain":{"domain":["direct.mit.edu"],"crossmark-restriction":true},"short-container-title":[],"published-print":{"date-parts":[[2013,4,1]]},"abstract":"<jats:title>Abstract<\/jats:title>\n               <jats:p>Learning algorithms need generally the ability to compare several streams of information. Neural learning architectures hence need a unit, a comparator, able to compare several inputs encoding either internal or external information, for instance, predictions and sensory readings. Without the possibility of comparing the values of predictions to actual sensory inputs, reward evaluation and supervised learning would not be possible.<\/jats:p>\n               <jats:p>Comparators are usually not implemented explicitly. Necessary comparisons are commonly performed by directly comparing the respective activities one-to-one. This implies that the characteristics of the two input streams (like size and encoding) must be provided at the time of designing the system. It is, however, plausible that biological comparators emerge from self-organizing, genetically encoded principles, which allow the system to adapt to the changes in the input and the organism. We propose an unsupervised neural circuitry, where the function of input comparison emerges via self-organization only from the interaction of the system with the respective inputs, without external influence or supervision.<\/jats:p>\n               <jats:p>The proposed neural comparator adapts in an unsupervised form according to the correlations present in the input streams. The system consists of a multilayer feedforward neural network, which follows a local output minimization (anti-Hebbian) rule for adaptation of the synaptic weights. The local output minimization allows the circuit to autonomously acquire the capability of comparing the neural activities received from different neural populations, which may differ in population size and the neural encoding used. The comparator is able to compare objects never encountered before in the sensory input streams and evaluate a measure of their similarity even when differently encoded.<\/jats:p>","DOI":"10.1162\/neco_a_00424","type":"journal-article","created":{"date-parts":[[2013,1,22]],"date-time":"2013-01-22T17:50:14Z","timestamp":1358877014000},"page":"1006-1028","update-policy":"http:\/\/dx.doi.org\/10.1162\/mitpressjournals.corrections.policy","source":"Crossref","is-referenced-by-count":3,"title":["A Self-Organized Neural Comparator"],"prefix":"10.1162","volume":"25","author":[{"given":"Guillermo A.","family":"Ludue\u00f1a","sequence":"first","affiliation":[{"name":"Institute for Theoretical Physics, Goethe University, Frankfurt am Main, Hessen 60438, Germany luduena@itp.uni-frankfurt.de"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"given":"Claudius","family":"Gros","sequence":"additional","affiliation":[{"name":"Institute for Theoretical Physics, Goethe University, Frankfurt am Main, Hessen 60438, Germany gros07@itp.uni-frankfurt.de"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"281","published-online":{"date-parts":[[2013,4,1]]},"reference":[{"key":"2022050215334068500_R1","doi-asserted-by":"crossref","unstructured":"Atiya, A. F.\n           (1990). An unsupervised learning technique for\n            artificial neural networks. Neural Networks,\n            3, 707\u2013711.","DOI":"10.1016\/0893-6080(90)90058-S"},{"key":"2022050215334068500_R2","doi-asserted-by":"crossref","unstructured":"Bell, A. J., &\n                Sejnowski,\n              T. (1995).\n            An information-maximization approach to blind separation and blind\n            deconvolution. Neural Computing, 7,\n            1129\u20131159.","DOI":"10.1162\/neco.1995.7.6.1129"},{"key":"2022050215334068500_R3","unstructured":"Billing, E. A.\n           (2010). Cognitive perspectives on robot\n            behavior. In Proceedings of the 2nd International Conference on\n            Agents and Artificial Intelligence: Special Session on Computing Languages with\n            Multi-Agent Systems and Bio-Inspired Devices (pp.\n            373\u2013382). Setubal,\n          Portugal: INSTICC."},{"key":"2022050215334068500_R4","unstructured":"Bovet, S., &\n                Pfeifer,\n            R. (2005a).\n            Emergence of coherent behaviors from homogeneous sensorimotor\n            coupling. In Proceedings of the 12th International Conference on\n            Advanced Robotics (ICAR). Piscataway, NJ:\n            IEEE."},{"key":"2022050215334068500_R5","doi-asserted-by":"crossref","unstructured":"Bovet, S., &\n                Pfeifer,\n            R. (2005b).\n            Emergence of delayed reward learning from sensorimotor\n            coordination. In Proceedings of the IEEE\/RSJ International\n            Conference on Intelligent Robots and Systems. Piscataway,\n            NJ: IEEE.","DOI":"10.1109\/IROS.2005.1545085"},{"key":"2022050215334068500_R6","doi-asserted-by":"crossref","unstructured":"Carpenter, G. A., &\n                Grossberg,\n              S. (1987).\n            A massively parallel architecture for a self-organizing neural pattern\n            recognition machine. Computer Vision, Graphics, and Image\n            Processing, 37,\n          54\u2013115.","DOI":"10.1016\/S0734-189X(87)80014-2"},{"key":"2022050215334068500_R7","doi-asserted-by":"crossref","unstructured":"Coultrip, R.,\n                Granger,\n            R., &\n              Lynch,\n            G. (1992).\n            A cortical model of winner-take-all competition via lateral\n            inhibition. Neural Networks, 5,\n            47\u201354.","DOI":"10.1016\/S0893-6080(05)80006-1"},{"key":"2022050215334068500_R8","doi-asserted-by":"crossref","unstructured":"Furao, S.,\n                Ogura, T.,\n            & Hasegawa,\n              O. (2007).\n            An enhanced self-organizing incremental neural network for online\n            unsupervised learning. Neural Networks,\n            20, 893\u2013903.","DOI":"10.1016\/j.neunet.2007.07.008"},{"key":"2022050215334068500_R9","doi-asserted-by":"crossref","unstructured":"Gros, C.\n           (2008). Complex and adaptive dynamical systems: A\n            primer. New York:\n            Springer.","DOI":"10.1007\/978-3-540-71874-1"},{"key":"2022050215334068500_R10","doi-asserted-by":"crossref","unstructured":"Guo, D.,\n                Shamai, S.,\n            & Verd\u00fa,\n            S. (2005).\n            Mutual information and minimum mean-square error in gaussian\n            channels. IEEE Transactions on Information Theory,\n            51, 1261\u20131282.","DOI":"10.1109\/TIT.2005.844072"},{"key":"2022050215334068500_R11","doi-asserted-by":"crossref","unstructured":"Japkowicz, N.\n           (2001). Supervised versus unsupervised\n            binary-learning by feedforward neural networks. Machine\n            Learning, 42,\n          97\u2013122.","DOI":"10.1023\/A:1007660820062"},{"key":"2022050215334068500_R12","unstructured":"Japkowicz, N.,\n                Myers, C.,\n            & Gluck,\n            M. (1995).\n            A novelty detection approach to classification. In\n            Proceedings of the Fourteenth Joint Conference on Artificial\n            Intelligence (pp. 518\u2013523).\n            Cambridge, MA: AAAI\n            Press."},{"key":"2022050215334068500_R13","doi-asserted-by":"crossref","unstructured":"Kaski, S., &\n                Kohonen,\n            T. (1994).\n            Winner-take-all networks for physiological models of competitive\n            learning. Neural Networks, 7,\n            973\u2013984.","DOI":"10.1016\/S0893-6080(05)80154-6"},{"key":"2022050215334068500_R14","doi-asserted-by":"crossref","unstructured":"Keller, J. M.,\n                Yager, R.\n              R., & Tahani,\n                W. (1992).\n            Neural network implementation of fuzzy logic.\n            Fuzzy Sets and Systems, 45,\n            1\u201312.","DOI":"10.1016\/0165-0114(92)90086-J"},{"key":"2022050215334068500_R15","doi-asserted-by":"crossref","unstructured":"Kohonen, T.\n           (1990). The self-organizing\n            map. Proceedings of the IEEE, 78,\n            1464\u20131480.","DOI":"10.1109\/5.58325"},{"key":"2022050215334068500_R16","doi-asserted-by":"crossref","unstructured":"Likhovidov, V.\n           (1997). Variational approach to unsupervised\n            learning algorithms of neural networks. Neural\n          Networks, 10,\n          273\u2013289.","DOI":"10.1016\/S0893-6080(96)00051-2"},{"key":"2022050215334068500_R17","doi-asserted-by":"crossref","unstructured":"Markovi\u0107, D., &\n                Gros,\n            C. (2010).\n            Self-organized chaos through polyhomeostatic\n          optimization. Physical Review Letters,\n            105, 068702.","DOI":"10.1103\/PhysRevLett.105.068702"},{"key":"2022050215334068500_R18","doi-asserted-by":"crossref","unstructured":"Markovi\u0107, D., &\n                Gros,\n            C. (2011).\n            Intrinsic adaptation in autonomous recurrent neural\n            networks. Neural Computation, 24,\n            523\u2013540.","DOI":"10.1162\/NECO_a_00232"},{"key":"2022050215334068500_R19","doi-asserted-by":"crossref","unstructured":"Mozzachiodi, R., &\n                Byrne, J.\n              H. (2010). More\n            than synaptic plasticity: Role of nonsynaptic plasticity in learning and\n            memory. Trends in Neurosciences, 33,\n            17\u201326.","DOI":"10.1016\/j.tins.2009.10.001"},{"key":"2022050215334068500_R20","doi-asserted-by":"crossref","unstructured":"O\u2019Reilly, R. C., &\n                Munakata,\n              Y. (2000).\n            Computational explorations in cognitive neuroscience.\n            Cambridge, MA: MIT\n          Press.","DOI":"10.7551\/mitpress\/2014.001.0001"},{"key":"2022050215334068500_R21","doi-asserted-by":"crossref","unstructured":"Sanger, T. D.\n           (1989). Optimal unsupervised learning in a\n            single-layer linear feedforward neural network. Neural\n            Networks, 2,\n          459\u2013473.","DOI":"10.1016\/0893-6080(89)90044-0"},{"key":"2022050215334068500_R22","doi-asserted-by":"crossref","unstructured":"Stemmler, M., &\n                Koch,\n            C. (1999).\n            How voltage-dependent conductances can adapt to maximize the information\n            encoded by neuronal firing rate. Nature Neuroscience,\n            2, 521\u2013527.","DOI":"10.1038\/9173"},{"key":"2022050215334068500_R23","doi-asserted-by":"crossref","unstructured":"Triesch, J.\n           (2005). A gradient rule for the plasticity of\n            a neuron\u2019s intrinsic excitability. In Proceedings of the\n            International Conference on Artificial Neural Networks (pp.\n            65\u201370). New York:\n            Springer.","DOI":"10.1007\/11550822_11"},{"key":"2022050215334068500_R24","doi-asserted-by":"crossref","unstructured":"Tong, H.,\n                Liu, T.,\n            & Tong,\n            Q. (2008).\n            Unsupervised learning neural network with convex constraint: Structure\n            and algorithm. Neurocomputing, 71,\n            620\u2013625.","DOI":"10.1016\/j.neucom.2007.03.016"}],"container-title":["Neural Computation"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/direct.mit.edu\/neco\/article-pdf\/25\/4\/1006\/2017471\/neco_a_00424.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/direct.mit.edu\/neco\/article-pdf\/25\/4\/1006\/2017471\/neco_a_00424.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2022,5,2]],"date-time":"2022-05-02T15:34:32Z","timestamp":1651505672000},"score":1,"resource":{"primary":{"URL":"https:\/\/direct.mit.edu\/neco\/article\/25\/4\/1006\/7883\/A-Self-Organized-Neural-Comparator"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2013,4,1]]},"references-count":24,"journal-issue":{"issue":"4","published-online":{"date-parts":[[2013,4,1]]},"published-print":{"date-parts":[[2013,4,1]]}},"URL":"https:\/\/doi.org\/10.1162\/neco_a_00424","relation":{},"ISSN":["0899-7667","1530-888X"],"issn-type":[{"value":"0899-7667","type":"print"},{"value":"1530-888X","type":"electronic"}],"subject":[],"published-other":{"date-parts":[[2013,4]]},"published":{"date-parts":[[2013,4,1]]}}}