Back propagation simulations using limited precision calculations
The precision required for neural net algorithms is an important question facing hardware architects. The authors present simulation results that compare floating point and limited precision integer back-propagation simulators. Data sets from the neural network benchmark suite maintained by Carnegie...
Gespeichert in:
| Hauptverfasser: | , |
|---|---|
| Format: | Tagungsbericht |
| Sprache: | eng |
| Schlagworte: | |
| Online-Zugang: | Volltext bestellen |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | 126 vol.2 |
|---|---|
| container_issue | |
| container_start_page | 121 |
| container_title | |
| container_volume | ii |
| creator | Holt, J.L. Baker, T.E. |
| description | The precision required for neural net algorithms is an important question facing hardware architects. The authors present simulation results that compare floating point and limited precision integer back-propagation simulators. Data sets from the neural network benchmark suite maintained by Carnegie Mellon University were used to compare integer and floating point implementations. The simulation results indicate that integer computation works quite well for the back-propagation algorithm. In all cases except one, the limited precision integer simulations performed as well as the floating point simulations. The effect of reducing the precision of the trained weights is also reported.< > |
| doi_str_mv | 10.1109/IJCNN.1991.155324 |
| format | Conference Proceeding |
| fullrecord | <record><control><sourceid>ieee_6IE</sourceid><recordid>TN_cdi_ieee_primary_155324</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><ieee_id>155324</ieee_id><sourcerecordid>155324</sourcerecordid><originalsourceid>FETCH-LOGICAL-c152t-c08e84aeef6b18327e072a7389348f9b1f8ad20de0d32653ad724eedc80800373</originalsourceid><addsrcrecordid>eNo1j81OwzAQhC0hJKD0AeCUF0jY9dqxcywRP0VVubTnyrU3lSFpozg98PYUCnP55vBppBHiDqFAhOph_lYvlwVWFRaoNUl1IW7AWCDAUuGVmKb0AacoDaW212L26Pxn1g-H3u3cGA_7LMXu2P7WlB1T3O-yNnZx5HCy2Mf043jX-n_pVlw2rk08_eNErJ-fVvVrvnh_mdezRe5RyzH3YNkqx9yUW7QkDYORzpCtSNmm2mJjXZAQGALJUpMLRirm4C1YADI0Effn3cjMm36InRu-NueT9A2U8kiT</addsrcrecordid><sourcetype>Publisher</sourcetype><iscdi>true</iscdi><recordtype>conference_proceeding</recordtype></control><display><type>conference_proceeding</type><title>Back propagation simulations using limited precision calculations</title><source>IEEE Electronic Library (IEL) Conference Proceedings</source><creator>Holt, J.L. ; Baker, T.E.</creator><creatorcontrib>Holt, J.L. ; Baker, T.E.</creatorcontrib><description>The precision required for neural net algorithms is an important question facing hardware architects. The authors present simulation results that compare floating point and limited precision integer back-propagation simulators. Data sets from the neural network benchmark suite maintained by Carnegie Mellon University were used to compare integer and floating point implementations. The simulation results indicate that integer computation works quite well for the back-propagation algorithm. In all cases except one, the limited precision integer simulations performed as well as the floating point simulations. The effect of reducing the precision of the trained weights is also reported.< ></description><identifier>ISBN: 0780301641</identifier><identifier>ISBN: 9780780301641</identifier><identifier>DOI: 10.1109/IJCNN.1991.155324</identifier><language>eng</language><publisher>IEEE</publisher><subject>Algorithm design and analysis ; Artificial neural networks ; Backpropagation algorithms ; Computational modeling ; Computer networks ; Convergence ; Neural network hardware ; Neural networks ; Testing</subject><ispartof>IJCNN-91-Seattle International Joint Conference on Neural Networks, 1991, Vol.ii, p.121-126 vol.2</ispartof><woscitedreferencessubscribed>false</woscitedreferencessubscribed><citedby>FETCH-LOGICAL-c152t-c08e84aeef6b18327e072a7389348f9b1f8ad20de0d32653ad724eedc80800373</citedby></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://ieeexplore.ieee.org/document/155324$$EHTML$$P50$$Gieee$$H</linktohtml><link.rule.ids>309,310,776,780,785,786,2052,4036,4037,27902,54895</link.rule.ids><linktorsrc>$$Uhttps://ieeexplore.ieee.org/document/155324$$EView_record_in_IEEE$$FView_record_in_$$GIEEE</linktorsrc></links><search><creatorcontrib>Holt, J.L.</creatorcontrib><creatorcontrib>Baker, T.E.</creatorcontrib><title>Back propagation simulations using limited precision calculations</title><title>IJCNN-91-Seattle International Joint Conference on Neural Networks</title><addtitle>IJCNN</addtitle><description>The precision required for neural net algorithms is an important question facing hardware architects. The authors present simulation results that compare floating point and limited precision integer back-propagation simulators. Data sets from the neural network benchmark suite maintained by Carnegie Mellon University were used to compare integer and floating point implementations. The simulation results indicate that integer computation works quite well for the back-propagation algorithm. In all cases except one, the limited precision integer simulations performed as well as the floating point simulations. The effect of reducing the precision of the trained weights is also reported.< ></description><subject>Algorithm design and analysis</subject><subject>Artificial neural networks</subject><subject>Backpropagation algorithms</subject><subject>Computational modeling</subject><subject>Computer networks</subject><subject>Convergence</subject><subject>Neural network hardware</subject><subject>Neural networks</subject><subject>Testing</subject><isbn>0780301641</isbn><isbn>9780780301641</isbn><fulltext>true</fulltext><rsrctype>conference_proceeding</rsrctype><creationdate>1991</creationdate><recordtype>conference_proceeding</recordtype><sourceid>6IE</sourceid><sourceid>RIE</sourceid><recordid>eNo1j81OwzAQhC0hJKD0AeCUF0jY9dqxcywRP0VVubTnyrU3lSFpozg98PYUCnP55vBppBHiDqFAhOph_lYvlwVWFRaoNUl1IW7AWCDAUuGVmKb0AacoDaW212L26Pxn1g-H3u3cGA_7LMXu2P7WlB1T3O-yNnZx5HCy2Mf043jX-n_pVlw2rk08_eNErJ-fVvVrvnh_mdezRe5RyzH3YNkqx9yUW7QkDYORzpCtSNmm2mJjXZAQGALJUpMLRirm4C1YADI0Effn3cjMm36InRu-NueT9A2U8kiT</recordid><startdate>1991</startdate><enddate>1991</enddate><creator>Holt, J.L.</creator><creator>Baker, T.E.</creator><general>IEEE</general><scope>6IE</scope><scope>6IL</scope><scope>CBEJK</scope><scope>RIE</scope><scope>RIL</scope></search><sort><creationdate>1991</creationdate><title>Back propagation simulations using limited precision calculations</title><author>Holt, J.L. ; Baker, T.E.</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-LOGICAL-c152t-c08e84aeef6b18327e072a7389348f9b1f8ad20de0d32653ad724eedc80800373</frbrgroupid><rsrctype>conference_proceedings</rsrctype><prefilter>conference_proceedings</prefilter><language>eng</language><creationdate>1991</creationdate><topic>Algorithm design and analysis</topic><topic>Artificial neural networks</topic><topic>Backpropagation algorithms</topic><topic>Computational modeling</topic><topic>Computer networks</topic><topic>Convergence</topic><topic>Neural network hardware</topic><topic>Neural networks</topic><topic>Testing</topic><toplevel>online_resources</toplevel><creatorcontrib>Holt, J.L.</creatorcontrib><creatorcontrib>Baker, T.E.</creatorcontrib><collection>IEEE Electronic Library (IEL) Conference Proceedings</collection><collection>IEEE Proceedings Order Plan All Online (POP All Online) 1998-present by volume</collection><collection>IEEE Xplore All Conference Proceedings</collection><collection>IEEE Xplore</collection><collection>IEEE Proceedings Order Plans (POP All) 1998-Present</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>Holt, J.L.</au><au>Baker, T.E.</au><format>book</format><genre>proceeding</genre><ristype>CONF</ristype><atitle>Back propagation simulations using limited precision calculations</atitle><btitle>IJCNN-91-Seattle International Joint Conference on Neural Networks</btitle><stitle>IJCNN</stitle><date>1991</date><risdate>1991</risdate><volume>ii</volume><spage>121</spage><epage>126 vol.2</epage><pages>121-126 vol.2</pages><isbn>0780301641</isbn><isbn>9780780301641</isbn><abstract>The precision required for neural net algorithms is an important question facing hardware architects. The authors present simulation results that compare floating point and limited precision integer back-propagation simulators. Data sets from the neural network benchmark suite maintained by Carnegie Mellon University were used to compare integer and floating point implementations. The simulation results indicate that integer computation works quite well for the back-propagation algorithm. In all cases except one, the limited precision integer simulations performed as well as the floating point simulations. The effect of reducing the precision of the trained weights is also reported.< ></abstract><pub>IEEE</pub><doi>10.1109/IJCNN.1991.155324</doi></addata></record> |
| fulltext | fulltext_linktorsrc |
| identifier | ISBN: 0780301641 |
| ispartof | IJCNN-91-Seattle International Joint Conference on Neural Networks, 1991, Vol.ii, p.121-126 vol.2 |
| issn | |
| language | eng |
| recordid | cdi_ieee_primary_155324 |
| source | IEEE Electronic Library (IEL) Conference Proceedings |
| subjects | Algorithm design and analysis Artificial neural networks Backpropagation algorithms Computational modeling Computer networks Convergence Neural network hardware Neural networks Testing |
| title | Back propagation simulations using limited precision calculations |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-02-01T09%3A48%3A43IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-ieee_6IE&rft_val_fmt=info:ofi/fmt:kev:mtx:book&rft.genre=proceeding&rft.atitle=Back%20propagation%20simulations%20using%20limited%20precision%20calculations&rft.btitle=IJCNN-91-Seattle%20International%20Joint%20Conference%20on%20Neural%20Networks&rft.au=Holt,%20J.L.&rft.date=1991&rft.volume=ii&rft.spage=121&rft.epage=126%20vol.2&rft.pages=121-126%20vol.2&rft.isbn=0780301641&rft.isbn_list=9780780301641&rft_id=info:doi/10.1109/IJCNN.1991.155324&rft_dat=%3Cieee_6IE%3E155324%3C/ieee_6IE%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rft_ieee_id=155324&rfr_iscdi=true |