Structural stability of unsupervised learning in feedback neural networks
Structural stability is proved for a large class of unsupervised nonlinear feedback neural networks, adaptive bidirectional associative memory (ABAM) models. The approach extends the ABAM models to the random-process domain as systems of stochastic differential equations and appends scaled Brownian...
Gespeichert in:
| Veröffentlicht in: | IEEE transactions on automatic control 1991-07, Vol.36 (7), p.785-792 |
|---|---|
| 1. Verfasser: | |
| Format: | Artikel |
| Sprache: | eng |
| Schlagworte: | |
| Online-Zugang: | Volltext bestellen |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | 792 |
|---|---|
| container_issue | 7 |
| container_start_page | 785 |
| container_title | IEEE transactions on automatic control |
| container_volume | 36 |
| creator | Kosko, B.A. |
| description | Structural stability is proved for a large class of unsupervised nonlinear feedback neural networks, adaptive bidirectional associative memory (ABAM) models. The approach extends the ABAM models to the random-process domain as systems of stochastic differential equations and appends scaled Brownian diffusions. It is also proved that this much larger family of models, random ABAM (RABAM) models, is globally stable. Intuitively, RABAM equilibria equal ABAM equilibria that vibrate randomly. The ABAM family includes many unsupervised feedback and feedforward neural models. All RABAM models permit Brownian annealing. The RABAM noise suppression theorem characterizes RABAM system vibration. The mean-squared activation and synaptic velocities decrease exponentially to their lower hounds, the respective temperature-scaled noise variances. The many neuronal and synaptic parameters missing from such neural network models are included, but as net random unmodeled effects. They do not affect the structure of real-time global computations.< > |
| doi_str_mv | 10.1109/9.85058 |
| format | Article |
| fullrecord | <record><control><sourceid>proquest_RIE</sourceid><recordid>TN_cdi_proquest_miscellaneous_28614830</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>85058</ieee_id><sourcerecordid>29052627</sourcerecordid><originalsourceid>FETCH-LOGICAL-c429t-6c46853383fc5d36f2bd629475cef556dba73e44c9a89740fa64cb01f0b5e9723</originalsourceid><addsrcrecordid>eNqF0DtPwzAQwHELgUQpiJktA4Ipxe_YI6p4VKrEAMyR45yRaeoUOwH125M-1LXTybqf_8MhdE3whBCsH_RECSzUCRoRIVROBWWnaIQxUbmmSp6ji5S-h6fknIzQ7L2Lve36aJosdabyje_WWeuyPqR-BfHXJ6izBkwMPnxlPmQOoK6MXWQBtr8CdH9tXKRLdOZMk-BqP8fo8_npY_qaz99eZtPHeW451V0uLZdKMKaYs6Jm0tGqllTzQlhwQsihXTDg3GqjdMGxM5LbChOHKwG6oGyM7nbdVWx_ekhdufTJQtOYAG2fSqqxoJIWx6GShCuGj0PBC4qLTfF-B21sU4rgylX0SxPXJcHl5vilLrfHH-TtPmmSNY2LJlifDlyQTVMM7GbHPAActrvEP4pzi0I</addsrcrecordid><sourcetype>Aggregation Database</sourcetype><iscdi>true</iscdi><recordtype>article</recordtype><pqid>25472077</pqid></control><display><type>article</type><title>Structural stability of unsupervised learning in feedback neural networks</title><source>IEEE Electronic Library (IEL)</source><creator>Kosko, B.A.</creator><creatorcontrib>Kosko, B.A.</creatorcontrib><description>Structural stability is proved for a large class of unsupervised nonlinear feedback neural networks, adaptive bidirectional associative memory (ABAM) models. The approach extends the ABAM models to the random-process domain as systems of stochastic differential equations and appends scaled Brownian diffusions. It is also proved that this much larger family of models, random ABAM (RABAM) models, is globally stable. Intuitively, RABAM equilibria equal ABAM equilibria that vibrate randomly. The ABAM family includes many unsupervised feedback and feedforward neural models. All RABAM models permit Brownian annealing. The RABAM noise suppression theorem characterizes RABAM system vibration. The mean-squared activation and synaptic velocities decrease exponentially to their lower hounds, the respective temperature-scaled noise variances. The many neuronal and synaptic parameters missing from such neural network models are included, but as net random unmodeled effects. They do not affect the structure of real-time global computations.< ></description><identifier>ISSN: 0018-9286</identifier><identifier>EISSN: 1558-2523</identifier><identifier>DOI: 10.1109/9.85058</identifier><identifier>CODEN: IETAA9</identifier><language>eng</language><publisher>New York, NY: IEEE</publisher><subject>Applied sciences ; Biological system modeling ; Calculus ; Differential equations ; Exact sciences and technology ; Information, signal and communications theory ; Intelligent networks ; Miscellaneous ; Neural networks ; Neurofeedback ; Signal processing ; Stability ; Stochastic processes ; Structural engineering ; Telecommunications and information theory ; Unsupervised learning</subject><ispartof>IEEE transactions on automatic control, 1991-07, Vol.36 (7), p.785-792</ispartof><rights>1992 INIST-CNRS</rights><lds50>peer_reviewed</lds50><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c429t-6c46853383fc5d36f2bd629475cef556dba73e44c9a89740fa64cb01f0b5e9723</citedby><cites>FETCH-LOGICAL-c429t-6c46853383fc5d36f2bd629475cef556dba73e44c9a89740fa64cb01f0b5e9723</cites></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/85058$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>314,780,784,796,27923,27924,54757</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/85058$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc><backlink>$$Uhttp://pascal-francis.inist.fr/vibad/index.php?action=getRecordDetail&idt=5125475$$DView record in Pascal Francis$$Hfree_for_read</backlink></links><search><creatorcontrib>Kosko, B.A.</creatorcontrib><title>Structural stability of unsupervised learning in feedback neural networks</title><title>IEEE transactions on automatic control</title><addtitle>TAC</addtitle><description>Structural stability is proved for a large class of unsupervised nonlinear feedback neural networks, adaptive bidirectional associative memory (ABAM) models. The approach extends the ABAM models to the random-process domain as systems of stochastic differential equations and appends scaled Brownian diffusions. It is also proved that this much larger family of models, random ABAM (RABAM) models, is globally stable. Intuitively, RABAM equilibria equal ABAM equilibria that vibrate randomly. The ABAM family includes many unsupervised feedback and feedforward neural models. All RABAM models permit Brownian annealing. The RABAM noise suppression theorem characterizes RABAM system vibration. The mean-squared activation and synaptic velocities decrease exponentially to their lower hounds, the respective temperature-scaled noise variances. The many neuronal and synaptic parameters missing from such neural network models are included, but as net random unmodeled effects. They do not affect the structure of real-time global computations.< ></description><subject>Applied sciences</subject><subject>Biological system modeling</subject><subject>Calculus</subject><subject>Differential equations</subject><subject>Exact sciences and technology</subject><subject>Information, signal and communications theory</subject><subject>Intelligent networks</subject><subject>Miscellaneous</subject><subject>Neural networks</subject><subject>Neurofeedback</subject><subject>Signal processing</subject><subject>Stability</subject><subject>Stochastic processes</subject><subject>Structural engineering</subject><subject>Telecommunications and information theory</subject><subject>Unsupervised learning</subject><issn>0018-9286</issn><issn>1558-2523</issn><fulltext>true</fulltext><rsrctype>article</rsrctype><creationdate>1991</creationdate><recordtype>article</recordtype><recordid>eNqF0DtPwzAQwHELgUQpiJktA4Ipxe_YI6p4VKrEAMyR45yRaeoUOwH125M-1LXTybqf_8MhdE3whBCsH_RECSzUCRoRIVROBWWnaIQxUbmmSp6ji5S-h6fknIzQ7L2Lve36aJosdabyje_WWeuyPqR-BfHXJ6izBkwMPnxlPmQOoK6MXWQBtr8CdH9tXKRLdOZMk-BqP8fo8_npY_qaz99eZtPHeW451V0uLZdKMKaYs6Jm0tGqllTzQlhwQsihXTDg3GqjdMGxM5LbChOHKwG6oGyM7nbdVWx_ekhdufTJQtOYAG2fSqqxoJIWx6GShCuGj0PBC4qLTfF-B21sU4rgylX0SxPXJcHl5vilLrfHH-TtPmmSNY2LJlifDlyQTVMM7GbHPAActrvEP4pzi0I</recordid><startdate>19910701</startdate><enddate>19910701</enddate><creator>Kosko, B.A.</creator><general>IEEE</general><general>Institute of Electrical and Electronics Engineers</general><scope>IQODW</scope><scope>AAYXX</scope><scope>CITATION</scope><scope>8FD</scope><scope>H8D</scope><scope>L7M</scope><scope>7SC</scope><scope>7SP</scope><scope>7TB</scope><scope>FR3</scope><scope>JQ2</scope><scope>L~C</scope><scope>L~D</scope></search><sort><creationdate>19910701</creationdate><title>Structural stability of unsupervised learning in feedback neural networks</title><author>Kosko, B.A.</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c429t-6c46853383fc5d36f2bd629475cef556dba73e44c9a89740fa64cb01f0b5e9723</frbrgroupid><rsrctype>articles</rsrctype><prefilter>articles</prefilter><language>eng</language><creationdate>1991</creationdate><topic>Applied sciences</topic><topic>Biological system modeling</topic><topic>Calculus</topic><topic>Differential equations</topic><topic>Exact sciences and technology</topic><topic>Information, signal and communications theory</topic><topic>Intelligent networks</topic><topic>Miscellaneous</topic><topic>Neural networks</topic><topic>Neurofeedback</topic><topic>Signal processing</topic><topic>Stability</topic><topic>Stochastic processes</topic><topic>Structural engineering</topic><topic>Telecommunications and information theory</topic><topic>Unsupervised learning</topic><toplevel>peer_reviewed</toplevel><toplevel>online_resources</toplevel><creatorcontrib>Kosko, B.A.</creatorcontrib><collection>Pascal-Francis</collection><collection>CrossRef</collection><collection>Technology Research Database</collection><collection>Aerospace Database</collection><collection>Advanced Technologies Database with Aerospace</collection><collection>Computer and Information Systems Abstracts</collection><collection>Electronics & Communications Abstracts</collection><collection>Mechanical & Transportation Engineering Abstracts</collection><collection>Engineering Research Database</collection><collection>ProQuest Computer Science Collection</collection><collection>Computer and Information Systems Abstracts Academic</collection><collection>Computer and Information Systems Abstracts Professional</collection><jtitle>IEEE transactions on automatic control</jtitle></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Kosko, B.A.</au><format>journal</format><genre>article</genre><ristype>JOUR</ristype><atitle>Structural stability of unsupervised learning in feedback neural networks</atitle><jtitle>IEEE transactions on automatic control</jtitle><stitle>TAC</stitle><date>1991-07-01</date><risdate>1991</risdate><volume>36</volume><issue>7</issue><spage>785</spage><epage>792</epage><pages>785-792</pages><issn>0018-9286</issn><eissn>1558-2523</eissn><coden>IETAA9</coden><abstract>Structural stability is proved for a large class of unsupervised nonlinear feedback neural networks, adaptive bidirectional associative memory (ABAM) models. The approach extends the ABAM models to the random-process domain as systems of stochastic differential equations and appends scaled Brownian diffusions. It is also proved that this much larger family of models, random ABAM (RABAM) models, is globally stable. Intuitively, RABAM equilibria equal ABAM equilibria that vibrate randomly. The ABAM family includes many unsupervised feedback and feedforward neural models. All RABAM models permit Brownian annealing. The RABAM noise suppression theorem characterizes RABAM system vibration. The mean-squared activation and synaptic velocities decrease exponentially to their lower hounds, the respective temperature-scaled noise variances. The many neuronal and synaptic parameters missing from such neural network models are included, but as net random unmodeled effects. They do not affect the structure of real-time global computations.< ></abstract><cop>New York, NY</cop><pub>IEEE</pub><doi>10.1109/9.85058</doi><tpages>8</tpages></addata></record> |
| fulltext | fulltext_linktorsrc |
| identifier | ISSN: 0018-9286 |
| ispartof | IEEE transactions on automatic control, 1991-07, Vol.36 (7), p.785-792 |
| issn | 0018-9286 1558-2523 |
| language | eng |
| recordid | cdi_proquest_miscellaneous_28614830 |
| source | IEEE Electronic Library (IEL) |
| subjects | Applied sciences Biological system modeling Calculus Differential equations Exact sciences and technology Information, signal and communications theory Intelligent networks Miscellaneous Neural networks Neurofeedback Signal processing Stability Stochastic processes Structural engineering Telecommunications and information theory Unsupervised learning |
| title | Structural stability of unsupervised learning in feedback neural networks |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-10T13%3A50%3A08IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-proquest_RIE&rft_val_fmt=info:ofi/fmt:kev:mtx:journal&rft.genre=article&rft.atitle=Structural%20stability%20of%20unsupervised%20learning%20in%20feedback%20neural%20networks&rft.jtitle=IEEE%20transactions%20on%20automatic%20control&rft.au=Kosko,%20B.A.&rft.date=1991-07-01&rft.volume=36&rft.issue=7&rft.spage=785&rft.epage=792&rft.pages=785-792&rft.issn=0018-9286&rft.eissn=1558-2523&rft.coden=IETAA9&rft_id=info:doi/10.1109/9.85058&rft_dat=%3Cproquest_RIE%3E29052627%3C/proquest_RIE%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_pqid=25472077&rft_id=info:pmid/&rft_ieee_id=85058&rfr_iscdi=true |