System and method for managing feature processing
The invention relates to a system and method for managing feature processing. At a machine learning service, a set of candidate variables is identified that can be used to train a model, including at least one processed variable generated by a feature processing transform. A cost estimate indicative...
Gespeichert in:
| Hauptverfasser: | , , , , , , , , , , , , , , , , , |
|---|---|
| Format: | Patent |
| Sprache: | chi ; eng |
| Schlagworte: | |
| Online-Zugang: | Volltext bestellen |
| Tags: |
Tag hinzufügen
Keine Tags, Fügen Sie den ersten Tag hinzu!
|
| container_end_page | |
|---|---|
| container_issue | |
| container_start_page | |
| container_title | |
| container_volume | |
| creator | INGERMAN ALEKSANDR MIKHAYLOVICH HERBRICH RALF KRISHNAN SRIRAM LI JIN ZHENG TIANMING BLICK DANIEL CORREA NICOLLE M DIRAC LEO PARKER DANNAKER CHARLES ERIC LEE POLLY PO YEE ZAHRATIAN, STEPHEN STEELE ROBERT MATTHIAS BRUCKNER MICHAEL RAMAKRISHNAN, RAMAKRISHNAN ZHUO DONGHUI AGARWAL TARUN PUVVADI SUDHAKAR RAO QIAN JUN |
| description | The invention relates to a system and method for managing feature processing. At a machine learning service, a set of candidate variables is identified that can be used to train a model, including at least one processed variable generated by a feature processing transform. A cost estimate indicative of an effect of implementing the feature processing transformation on a performance metric associated with a predicted goal of the model is determined. Based at least in part on the cost estimate, a feature processing proposal is implemented that excludes the feature processing transform.
本公开涉及用于管理特征处理的系统和方法。在机器学习服务处，识别可用于训练模型的一组候选变量，包括通过特征处理变换产生的至少一个处理的变量。确定指示实现所述特征处理变换对与所述模型的预测目标相关联的性能度量的影响的成本估计。至少部分地基于所述成本估计，实现排除所述特征处理变换的特征处理提议。 |
| format | Patent |
| fullrecord | <record><control><sourceid>epo_EVB</sourceid><recordid>TN_cdi_epo_espacenet_CN118152127A</recordid><sourceformat>XML</sourceformat><sourcesystem>PC</sourcesystem><sourcerecordid>CN118152127A</sourcerecordid><originalsourceid>FETCH-epo_espacenet_CN118152127A3</originalsourceid><addsrcrecordid>eNrjZDAMriwuSc1VSMxLUchNLcnIT1FIyy9SyE3MS0zPzEtXSEtNLCktSlUoKMpPTi0uBgrxMLCmJeYUp_JCaW4GRTfXEGcP3dSC_PjU4oLE5NS81JJ4Zz9DQwtDUyNDI3NHY2LUAACX-yu7</addsrcrecordid><sourcetype>Open Access Repository</sourcetype><iscdi>true</iscdi><recordtype>patent</recordtype></control><display><type>patent</type><title>System and method for managing feature processing</title><source>esp@cenet</source><creator>INGERMAN ALEKSANDR MIKHAYLOVICH ; HERBRICH RALF ; KRISHNAN SRIRAM ; LI JIN ; ZHENG TIANMING ; BLICK DANIEL ; CORREA NICOLLE M ; DIRAC LEO PARKER ; DANNAKER CHARLES ERIC ; LEE POLLY PO YEE ; ZAHRATIAN, STEPHEN ; STEELE ROBERT MATTHIAS ; BRUCKNER MICHAEL ; RAMAKRISHNAN, RAMAKRISHNAN ; ZHUO DONGHUI ; AGARWAL TARUN ; PUVVADI SUDHAKAR RAO ; QIAN JUN</creator><creatorcontrib>INGERMAN ALEKSANDR MIKHAYLOVICH ; HERBRICH RALF ; KRISHNAN SRIRAM ; LI JIN ; ZHENG TIANMING ; BLICK DANIEL ; CORREA NICOLLE M ; DIRAC LEO PARKER ; DANNAKER CHARLES ERIC ; LEE POLLY PO YEE ; ZAHRATIAN, STEPHEN ; STEELE ROBERT MATTHIAS ; BRUCKNER MICHAEL ; RAMAKRISHNAN, RAMAKRISHNAN ; ZHUO DONGHUI ; AGARWAL TARUN ; PUVVADI SUDHAKAR RAO ; QIAN JUN</creatorcontrib><description>The invention relates to a system and method for managing feature processing. At a machine learning service, a set of candidate variables is identified that can be used to train a model, including at least one processed variable generated by a feature processing transform. A cost estimate indicative of an effect of implementing the feature processing transformation on a performance metric associated with a predicted goal of the model is determined. Based at least in part on the cost estimate, a feature processing proposal is implemented that excludes the feature processing transform.
本公开涉及用于管理特征处理的系统和方法。在机器学习服务处，识别可用于训练模型的一组候选变量，包括通过特征处理变换产生的至少一个处理的变量。确定指示实现所述特征处理变换对与所述模型的预测目标相关联的性能度量的影响的成本估计。至少部分地基于所述成本估计，实现排除所述特征处理变换的特征处理提议。</description><language>chi ; eng</language><subject>CALCULATING ; COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS ; COMPUTING ; COUNTING ; ELECTRIC DIGITAL DATA PROCESSING ; PHYSICS</subject><creationdate>2024</creationdate><oa>free_for_read</oa><woscitedreferencessubscribed>false</woscitedreferencessubscribed></display><links><openurl>$$Topenurl_article</openurl><openurlfulltext>$$Topenurlfull_article</openurlfulltext><thumbnail>$$Tsyndetics_thumb_exl</thumbnail><linktohtml>$$Uhttps://worldwide.espacenet.com/publicationDetails/biblio?FT=D&date=20240607&DB=EPODOC&CC=CN&NR=118152127A$$EHTML$$P50$$Gepo$$Hfree_for_read</linktohtml><link.rule.ids>230,308,780,885,25564,76547</link.rule.ids><linktorsrc>$$Uhttps://worldwide.espacenet.com/publicationDetails/biblio?FT=D&date=20240607&DB=EPODOC&CC=CN&NR=118152127A$$EView_record_in_European_Patent_Office$$FView_record_in_$$GEuropean_Patent_Office$$Hfree_for_read</linktorsrc></links><search><creatorcontrib>INGERMAN ALEKSANDR MIKHAYLOVICH</creatorcontrib><creatorcontrib>HERBRICH RALF</creatorcontrib><creatorcontrib>KRISHNAN SRIRAM</creatorcontrib><creatorcontrib>LI JIN</creatorcontrib><creatorcontrib>ZHENG TIANMING</creatorcontrib><creatorcontrib>BLICK DANIEL</creatorcontrib><creatorcontrib>CORREA NICOLLE M</creatorcontrib><creatorcontrib>DIRAC LEO PARKER</creatorcontrib><creatorcontrib>DANNAKER CHARLES ERIC</creatorcontrib><creatorcontrib>LEE POLLY PO YEE</creatorcontrib><creatorcontrib>ZAHRATIAN, STEPHEN</creatorcontrib><creatorcontrib>STEELE ROBERT MATTHIAS</creatorcontrib><creatorcontrib>BRUCKNER MICHAEL</creatorcontrib><creatorcontrib>RAMAKRISHNAN, RAMAKRISHNAN</creatorcontrib><creatorcontrib>ZHUO DONGHUI</creatorcontrib><creatorcontrib>AGARWAL TARUN</creatorcontrib><creatorcontrib>PUVVADI SUDHAKAR RAO</creatorcontrib><creatorcontrib>QIAN JUN</creatorcontrib><title>System and method for managing feature processing</title><description>The invention relates to a system and method for managing feature processing. At a machine learning service, a set of candidate variables is identified that can be used to train a model, including at least one processed variable generated by a feature processing transform. A cost estimate indicative of an effect of implementing the feature processing transformation on a performance metric associated with a predicted goal of the model is determined. Based at least in part on the cost estimate, a feature processing proposal is implemented that excludes the feature processing transform.
本公开涉及用于管理特征处理的系统和方法。在机器学习服务处，识别可用于训练模型的一组候选变量，包括通过特征处理变换产生的至少一个处理的变量。确定指示实现所述特征处理变换对与所述模型的预测目标相关联的性能度量的影响的成本估计。至少部分地基于所述成本估计，实现排除所述特征处理变换的特征处理提议。</description><subject>CALCULATING</subject><subject>COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS</subject><subject>COMPUTING</subject><subject>COUNTING</subject><subject>ELECTRIC DIGITAL DATA PROCESSING</subject><subject>PHYSICS</subject><fulltext>true</fulltext><rsrctype>patent</rsrctype><creationdate>2024</creationdate><recordtype>patent</recordtype><sourceid>EVB</sourceid><recordid>eNrjZDAMriwuSc1VSMxLUchNLcnIT1FIyy9SyE3MS0zPzEtXSEtNLCktSlUoKMpPTi0uBgrxMLCmJeYUp_JCaW4GRTfXEGcP3dSC_PjU4oLE5NS81JJ4Zz9DQwtDUyNDI3NHY2LUAACX-yu7</recordid><startdate>20240607</startdate><enddate>20240607</enddate><creator>INGERMAN ALEKSANDR MIKHAYLOVICH</creator><creator>HERBRICH RALF</creator><creator>KRISHNAN SRIRAM</creator><creator>LI JIN</creator><creator>ZHENG TIANMING</creator><creator>BLICK DANIEL</creator><creator>CORREA NICOLLE M</creator><creator>DIRAC LEO PARKER</creator><creator>DANNAKER CHARLES ERIC</creator><creator>LEE POLLY PO YEE</creator><creator>ZAHRATIAN, STEPHEN</creator><creator>STEELE ROBERT MATTHIAS</creator><creator>BRUCKNER MICHAEL</creator><creator>RAMAKRISHNAN, RAMAKRISHNAN</creator><creator>ZHUO DONGHUI</creator><creator>AGARWAL TARUN</creator><creator>PUVVADI SUDHAKAR RAO</creator><creator>QIAN JUN</creator><scope>EVB</scope></search><sort><creationdate>20240607</creationdate><title>System and method for managing feature processing</title><author>INGERMAN ALEKSANDR MIKHAYLOVICH ; HERBRICH RALF ; KRISHNAN SRIRAM ; LI JIN ; ZHENG TIANMING ; BLICK DANIEL ; CORREA NICOLLE M ; DIRAC LEO PARKER ; DANNAKER CHARLES ERIC ; LEE POLLY PO YEE ; ZAHRATIAN, STEPHEN ; STEELE ROBERT MATTHIAS ; BRUCKNER MICHAEL ; RAMAKRISHNAN, RAMAKRISHNAN ; ZHUO DONGHUI ; AGARWAL TARUN ; PUVVADI SUDHAKAR RAO ; QIAN JUN</author></sort><facets><frbrtype>5</frbrtype><frbrgroupid>cdi_FETCH-epo_espacenet_CN118152127A3</frbrgroupid><rsrctype>patents</rsrctype><prefilter>patents</prefilter><language>chi ; eng</language><creationdate>2024</creationdate><topic>CALCULATING</topic><topic>COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS</topic><topic>COMPUTING</topic><topic>COUNTING</topic><topic>ELECTRIC DIGITAL DATA PROCESSING</topic><topic>PHYSICS</topic><toplevel>online_resources</toplevel><creatorcontrib>INGERMAN ALEKSANDR MIKHAYLOVICH</creatorcontrib><creatorcontrib>HERBRICH RALF</creatorcontrib><creatorcontrib>KRISHNAN SRIRAM</creatorcontrib><creatorcontrib>LI JIN</creatorcontrib><creatorcontrib>ZHENG TIANMING</creatorcontrib><creatorcontrib>BLICK DANIEL</creatorcontrib><creatorcontrib>CORREA NICOLLE M</creatorcontrib><creatorcontrib>DIRAC LEO PARKER</creatorcontrib><creatorcontrib>DANNAKER CHARLES ERIC</creatorcontrib><creatorcontrib>LEE POLLY PO YEE</creatorcontrib><creatorcontrib>ZAHRATIAN, STEPHEN</creatorcontrib><creatorcontrib>STEELE ROBERT MATTHIAS</creatorcontrib><creatorcontrib>BRUCKNER MICHAEL</creatorcontrib><creatorcontrib>RAMAKRISHNAN, RAMAKRISHNAN</creatorcontrib><creatorcontrib>ZHUO DONGHUI</creatorcontrib><creatorcontrib>AGARWAL TARUN</creatorcontrib><creatorcontrib>PUVVADI SUDHAKAR RAO</creatorcontrib><creatorcontrib>QIAN JUN</creatorcontrib><collection>esp@cenet</collection></facets><delivery><delcategory>Remote Search Resource</delcategory><fulltext>fulltext_linktorsrc</fulltext></delivery><addata><au>INGERMAN ALEKSANDR MIKHAYLOVICH</au><au>HERBRICH RALF</au><au>KRISHNAN SRIRAM</au><au>LI JIN</au><au>ZHENG TIANMING</au><au>BLICK DANIEL</au><au>CORREA NICOLLE M</au><au>DIRAC LEO PARKER</au><au>DANNAKER CHARLES ERIC</au><au>LEE POLLY PO YEE</au><au>ZAHRATIAN, STEPHEN</au><au>STEELE ROBERT MATTHIAS</au><au>BRUCKNER MICHAEL</au><au>RAMAKRISHNAN, RAMAKRISHNAN</au><au>ZHUO DONGHUI</au><au>AGARWAL TARUN</au><au>PUVVADI SUDHAKAR RAO</au><au>QIAN JUN</au><format>patent</format><genre>patent</genre><ristype>GEN</ristype><title>System and method for managing feature processing</title><date>2024-06-07</date><risdate>2024</risdate><abstract>The invention relates to a system and method for managing feature processing. At a machine learning service, a set of candidate variables is identified that can be used to train a model, including at least one processed variable generated by a feature processing transform. A cost estimate indicative of an effect of implementing the feature processing transformation on a performance metric associated with a predicted goal of the model is determined. Based at least in part on the cost estimate, a feature processing proposal is implemented that excludes the feature processing transform.
本公开涉及用于管理特征处理的系统和方法。在机器学习服务处，识别可用于训练模型的一组候选变量，包括通过特征处理变换产生的至少一个处理的变量。确定指示实现所述特征处理变换对与所述模型的预测目标相关联的性能度量的影响的成本估计。至少部分地基于所述成本估计，实现排除所述特征处理变换的特征处理提议。</abstract><oa>free_for_read</oa></addata></record> |
| fulltext | fulltext_linktorsrc |
| identifier | |
| ispartof | |
| issn | |
| language | chi ; eng |
| recordid | cdi_epo_espacenet_CN118152127A |
| source | esp@cenet |
| subjects | CALCULATING COMPUTER SYSTEMS BASED ON SPECIFIC COMPUTATIONAL MODELS COMPUTING COUNTING ELECTRIC DIGITAL DATA PROCESSING PHYSICS |
| title | System and method for managing feature processing |
| url | https://sfx.bib-bvb.de/sfx_tum?ctx_ver=Z39.88-2004&ctx_enc=info:ofi/enc:UTF-8&ctx_tim=2025-01-07T10%3A54%3A12IST&url_ver=Z39.88-2004&url_ctx_fmt=infofi/fmt:kev:mtx:ctx&rfr_id=info:sid/primo.exlibrisgroup.com:primo3-Article-epo_EVB&rft_val_fmt=info:ofi/fmt:kev:mtx:patent&rft.genre=patent&rft.au=INGERMAN%20ALEKSANDR%20MIKHAYLOVICH&rft.date=2024-06-07&rft_id=info:doi/&rft_dat=%3Cepo_EVB%3ECN118152127A%3C/epo_EVB%3E%3Curl%3E%3C/url%3E&disable_directlink=true&sfx.directlink=off&sfx.report_link=0&rft_id=info:oai/&rft_id=info:pmid/&rfr_iscdi=true |