Self-certifying classification by linearized deep assignment
We propose a novel class of deep stochastic predictors for classifying metric data on graphs within the PAC-Bayes risk certification paradigm. Classifiers are realized as linearly parametrized deep assignment flows with random initial conditions. Building on the recent PAC-Bayes literature and data-...
Gespeichert in:
| Hauptverfasser: | , , , |
|---|---|
| Dokumenttyp: | Article (Journal) Kapitel/Artikel |
| Sprache: | Englisch |
| Veröffentlicht: |
18 Feb 2022
|
| Ausgabe: | Version v2 |
| In: |
Arxiv
Year: 2022, Pages: 1-19 |
| DOI: | 10.48550/arXiv.2201.11162 |
| Online-Zugang: | Verlag, kostenfrei, Volltext: https://doi.org/10.48550/arXiv.2201.11162 Verlag, kostenfrei, Volltext: http://arxiv.org/abs/2201.11162 |
| Verfasserangaben: | Bastian Boll, Alexander Zeilmann, Stefania Petra, Christoph Schnörr |
MARC
| LEADER | 00000caa a2200000 c 4500 | ||
|---|---|---|---|
| 001 | 1818764601 | ||
| 003 | DE-627 | ||
| 005 | 20240110143340.0 | ||
| 007 | cr uuu---uuuuu | ||
| 008 | 221012s2022 xx |||||o 00| ||eng c | ||
| 024 | 7 | |a 10.48550/arXiv.2201.11162 |2 doi | |
| 035 | |a (DE-627)1818764601 | ||
| 035 | |a (DE-599)KXP1818764601 | ||
| 035 | |a (OCoLC)1361695775 | ||
| 040 | |a DE-627 |b ger |c DE-627 |e rda | ||
| 041 | |a eng | ||
| 084 | |a 27 |2 sdnb | ||
| 100 | 1 | |a Boll, Bastian |d 1996- |e VerfasserIn |0 (DE-588)1241416443 |0 (DE-627)1770926682 |4 aut | |
| 245 | 1 | 0 | |a Self-certifying classification by linearized deep assignment |c Bastian Boll, Alexander Zeilmann, Stefania Petra, Christoph Schnörr |
| 250 | |a Version v2 | ||
| 264 | 1 | |c 18 Feb 2022 | |
| 300 | |a 19 | ||
| 336 | |a Text |b txt |2 rdacontent | ||
| 337 | |a Computermedien |b c |2 rdamedia | ||
| 338 | |a Online-Ressource |b cr |2 rdacarrier | ||
| 500 | |a Online veröffentlicht am 26. Januar 2022 | ||
| 500 | |a Gesehen am 10.01.2024 | ||
| 520 | |a We propose a novel class of deep stochastic predictors for classifying metric data on graphs within the PAC-Bayes risk certification paradigm. Classifiers are realized as linearly parametrized deep assignment flows with random initial conditions. Building on the recent PAC-Bayes literature and data-dependent priors, this approach enables (i) to use risk bounds as training objectives for learning posterior distributions on the hypothesis space and (ii) to compute tight out-of-sample risk certificates of randomized classifiers more efficiently than related work. Comparison with empirical test set errors illustrates the performance and practicality of this self-certifying classification method. | ||
| 650 | 4 | |a Computer Science - Machine Learning | |
| 650 | 4 | |a Mathematics - Optimization and Control | |
| 650 | 4 | |a Statistics - Machine Learning | |
| 700 | 1 | |a Zeilmann, Alexander |e VerfasserIn |0 (DE-588)1225318092 |0 (DE-627)1744906696 |4 aut | |
| 700 | 1 | |a Petra, Stefania |e VerfasserIn |0 (DE-588)1065905580 |0 (DE-627)816924961 |0 (DE-576)425560155 |4 aut | |
| 700 | 1 | |a Schnörr, Christoph |e VerfasserIn |0 (DE-588)1023033348 |0 (DE-627)717351017 |0 (DE-576)168404540 |4 aut | |
| 773 | 0 | 8 | |i Enthalten in |t Arxiv |d Ithaca, NY : Cornell University, 1991 |g (2022), Artikel-ID 2201.11162, Seite 1-19 |h Online-Ressource |w (DE-627)509006531 |w (DE-600)2225896-6 |w (DE-576)28130436X |7 nnas |a Self-certifying classification by linearized deep assignment |
| 773 | 1 | 8 | |g year:2022 |g elocationid:2201.11162 |g pages:1-19 |g extent:19 |a Self-certifying classification by linearized deep assignment |
| 856 | 4 | 0 | |u https://doi.org/10.48550/arXiv.2201.11162 |x Verlag |x Resolving-System |z kostenfrei |3 Volltext |
| 856 | 4 | 0 | |u http://arxiv.org/abs/2201.11162 |x Verlag |z kostenfrei |3 Volltext |
| 951 | |a AR | ||
| 992 | |a 20221012 | ||
| 993 | |a Article | ||
| 994 | |a 2022 | ||
| 998 | |g 1023033348 |a Schnörr, Christoph |m 1023033348:Schnörr, Christoph |d 110000 |d 110200 |d 110000 |d 110400 |d 700000 |d 728500 |e 110000PS1023033348 |e 110200PS1023033348 |e 110000PS1023033348 |e 110400PS1023033348 |e 700000PS1023033348 |e 728500PS1023033348 |k 0/110000/ |k 1/110000/110200/ |k 0/110000/ |k 1/110000/110400/ |k 0/700000/ |k 1/700000/728500/ |p 4 |y j | ||
| 998 | |g 1065905580 |a Petra, Stefania |m 1065905580:Petra, Stefania |d 110000 |d 110200 |d 110000 |d 110400 |e 110000PP1065905580 |e 110200PP1065905580 |e 110000PP1065905580 |e 110400PP1065905580 |k 0/110000/ |k 1/110000/110200/ |k 0/110000/ |k 1/110000/110400/ |p 3 | ||
| 998 | |g 1225318092 |a Zeilmann, Alexander |m 1225318092:Zeilmann, Alexander |d 110000 |d 110200 |d 110000 |d 110400 |d 700000 |d 728500 |e 110000PZ1225318092 |e 110200PZ1225318092 |e 110000PZ1225318092 |e 110400PZ1225318092 |e 700000PZ1225318092 |e 728500PZ1225318092 |k 0/110000/ |k 1/110000/110200/ |k 0/110000/ |k 1/110000/110400/ |k 0/700000/ |k 1/700000/728500/ |p 2 | ||
| 998 | |g 1241416443 |a Boll, Bastian |m 1241416443:Boll, Bastian |d 110000 |d 110200 |d 110000 |d 110400 |e 110000PB1241416443 |e 110200PB1241416443 |e 110000PB1241416443 |e 110400PB1241416443 |k 0/110000/ |k 1/110000/110200/ |k 0/110000/ |k 1/110000/110400/ |p 1 |x j | ||
| 999 | |a KXP-PPN1818764601 |e 4196392091 | ||
| BIB | |a Y | ||
| JSO | |a {"language":["eng"],"recId":"1818764601","type":{"bibl":"chapter","media":"Online-Ressource"},"note":["Online veröffentlicht am 26. Januar 2022","Gesehen am 10.01.2024"],"title":[{"title_sort":"Self-certifying classification by linearized deep assignment","title":"Self-certifying classification by linearized deep assignment"}],"person":[{"role":"aut","display":"Boll, Bastian","roleDisplay":"VerfasserIn","given":"Bastian","family":"Boll"},{"roleDisplay":"VerfasserIn","display":"Zeilmann, Alexander","role":"aut","family":"Zeilmann","given":"Alexander"},{"given":"Stefania","family":"Petra","role":"aut","roleDisplay":"VerfasserIn","display":"Petra, Stefania"},{"display":"Schnörr, Christoph","roleDisplay":"VerfasserIn","role":"aut","family":"Schnörr","given":"Christoph"}],"relHost":[{"recId":"509006531","language":["eng"],"note":["Gesehen am 28.05.2024"],"disp":"Self-certifying classification by linearized deep assignmentArxiv","type":{"media":"Online-Ressource","bibl":"edited-book"},"titleAlt":[{"title":"Arxiv.org"},{"title":"Arxiv.org e-print archive"},{"title":"Arxiv e-print archive"},{"title":"De.arxiv.org"}],"part":{"pages":"1-19","year":"2022","extent":"19","text":"(2022), Artikel-ID 2201.11162, Seite 1-19"},"pubHistory":["1991 -"],"title":[{"title":"Arxiv","title_sort":"Arxiv"}],"physDesc":[{"extent":"Online-Ressource"}],"id":{"zdb":["2225896-6"],"eki":["509006531"]},"origin":[{"publisherPlace":"Ithaca, NY ; [Erscheinungsort nicht ermittelbar]","dateIssuedDisp":"1991-","publisher":"Cornell University ; Arxiv.org","dateIssuedKey":"1991"}]}],"physDesc":[{"extent":"19 S."}],"id":{"eki":["1818764601"],"doi":["10.48550/arXiv.2201.11162"]},"origin":[{"dateIssuedKey":"2022","dateIssuedDisp":"18 Feb 2022","edition":"Version v2"}],"name":{"displayForm":["Bastian Boll, Alexander Zeilmann, Stefania Petra, Christoph Schnörr"]}} | ||
| SRT | |a BOLLBASTIASELFCERTIF1820 | ||