Beyond Point Estimate: Inferring Ensemble Prediction Variation from Neuron Activation Strength in Recommender Systems
Despite deep neural network (DNN)'s impressive prediction performance in various domains, it is well known now that a set of DNN models trained with the same model specification and the same data can produce very different prediction results. Ensemble method is one state-of-the-art benchmark fo...
Saved in:
Published in | arXiv.org |
---|---|
Main Authors | , , , , , , |
Format | Paper |
Language | English |
Published |
Ithaca
Cornell University Library, arXiv.org
17.08.2020
|
Subjects | |
Online Access | Get full text |
Cover
Loading…
Abstract | Despite deep neural network (DNN)'s impressive prediction performance in various domains, it is well known now that a set of DNN models trained with the same model specification and the same data can produce very different prediction results. Ensemble method is one state-of-the-art benchmark for prediction uncertainty estimation. However, ensembles are expensive to train and serve for web-scale traffic. In this paper, we seek to advance the understanding of prediction variation estimated by the ensemble method. Through empirical experiments on two widely used benchmark datasets MovieLens and Criteo in recommender systems, we observe that prediction variations come from various randomness sources, including training data shuffling, and parameter random initialization. By introducing more randomness into model training, we notice that ensemble's mean predictions tend to be more accurate while the prediction variations tend to be higher. Moreover, we propose to infer prediction variation from neuron activation strength and demonstrate the strong prediction power from activation strength features. Our experiment results show that the average R squared on MovieLens is as high as 0.56 and on Criteo is 0.81. Our method performs especially well when detecting the lowest and highest variation buckets, with 0.92 AUC and 0.89 AUC respectively. Our approach provides a simple way for prediction variation estimation, which opens up new opportunities for future work in many interesting areas (e.g.,model-based reinforcement learning) without relying on serving expensive ensemble models. |
---|---|
AbstractList | Despite deep neural network (DNN)'s impressive prediction performance in various domains, it is well known now that a set of DNN models trained with the same model specification and the same data can produce very different prediction results. Ensemble method is one state-of-the-art benchmark for prediction uncertainty estimation. However, ensembles are expensive to train and serve for web-scale traffic. In this paper, we seek to advance the understanding of prediction variation estimated by the ensemble method. Through empirical experiments on two widely used benchmark datasets MovieLens and Criteo in recommender systems, we observe that prediction variations come from various randomness sources, including training data shuffling, and parameter random initialization. By introducing more randomness into model training, we notice that ensemble's mean predictions tend to be more accurate while the prediction variations tend to be higher. Moreover, we propose to infer prediction variation from neuron activation strength and demonstrate the strong prediction power from activation strength features. Our experiment results show that the average R squared on MovieLens is as high as 0.56 and on Criteo is 0.81. Our method performs especially well when detecting the lowest and highest variation buckets, with 0.92 AUC and 0.89 AUC respectively. Our approach provides a simple way for prediction variation estimation, which opens up new opportunities for future work in many interesting areas (e.g.,model-based reinforcement learning) without relying on serving expensive ensemble models. |
Author | Cui, Claire Wang, Yuyan Derek Zhiyuan Cheng Chi, Ed H Lin, Dong Hong, Lichan Chen, Zhe |
Author_xml | – sequence: 1 givenname: Zhe surname: Chen fullname: Chen, Zhe – sequence: 2 givenname: Yuyan surname: Wang fullname: Wang, Yuyan – sequence: 3 givenname: Dong surname: Lin fullname: Lin, Dong – sequence: 4 fullname: Derek Zhiyuan Cheng – sequence: 5 givenname: Lichan surname: Hong fullname: Hong, Lichan – sequence: 6 givenname: Ed surname: Chi middlename: H fullname: Chi, Ed H – sequence: 7 givenname: Claire surname: Cui fullname: Cui, Claire |
BookMark | eNqNis0OATEURhsh8fsON7GWVGsQO2SEjYgRWxnmDhW95bYj8fYmeACr7-Q7pymq5AgroqG07vfGA6XqouP9VUqphiMVRbohihm-HGWwcYYCxD4YmwacwIpyZDZ0hpg82uMNYcOYmVMwjmCfskk_lLOzsMaCS56W8vm9k8BI53ABQ7DFk7MWKUOG5OUDWt8WtTy9eez8tiW6i3g3X_bu7B4F-nC4uoKpVAc10JGU4_5oqP-r3ksWTto |
ContentType | Paper |
Copyright | 2020. This work is published under http://creativecommons.org/licenses/by-nc-sa/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License. |
Copyright_xml | – notice: 2020. This work is published under http://creativecommons.org/licenses/by-nc-sa/4.0/ (the “License”). Notwithstanding the ProQuest Terms and Conditions, you may use this content in accordance with the terms of the License. |
DBID | 8FE 8FG ABJCF ABUWG AFKRA AZQEC BENPR BGLVJ CCPQU DWQXO HCIFZ L6V M7S PIMPY PQEST PQQKQ PQUKI PRINS PTHSS |
DatabaseName | ProQuest SciTech Collection ProQuest Technology Collection Materials Science & Engineering Collection ProQuest Central (Alumni) ProQuest Central ProQuest Central Essentials ProQuest Central Technology Collection ProQuest One Community College ProQuest Central Korea SciTech Premium Collection ProQuest Engineering Collection Engineering Database Publicly Available Content Database ProQuest One Academic Eastern Edition (DO NOT USE) ProQuest One Academic ProQuest One Academic UKI Edition ProQuest Central China Engineering Collection |
DatabaseTitle | Publicly Available Content Database Engineering Database Technology Collection ProQuest Central Essentials ProQuest One Academic Eastern Edition ProQuest Central (Alumni Edition) SciTech Premium Collection ProQuest One Community College ProQuest Technology Collection ProQuest SciTech Collection ProQuest Central China ProQuest Central ProQuest Engineering Collection ProQuest One Academic UKI Edition ProQuest Central Korea Materials Science & Engineering Collection ProQuest One Academic Engineering Collection |
DatabaseTitleList | Publicly Available Content Database |
Database_xml | – sequence: 1 dbid: 8FG name: ProQuest Technology Collection url: https://search.proquest.com/technologycollection1 sourceTypes: Aggregation Database |
DeliveryMethod | fulltext_linktorsrc |
Discipline | Physics |
EISSN | 2331-8422 |
Genre | Working Paper/Pre-Print |
GroupedDBID | 8FE 8FG ABJCF ABUWG AFKRA ALMA_UNASSIGNED_HOLDINGS AZQEC BENPR BGLVJ CCPQU DWQXO FRJ HCIFZ L6V M7S M~E PIMPY PQEST PQQKQ PQUKI PRINS PTHSS |
ID | FETCH-proquest_journals_24350081763 |
IEDL.DBID | 8FG |
IngestDate | Thu Oct 10 18:03:28 EDT 2024 |
IsOpenAccess | true |
IsPeerReviewed | false |
IsScholarly | false |
Language | English |
LinkModel | DirectLink |
MergedId | FETCHMERGED-proquest_journals_24350081763 |
OpenAccessLink | https://www.proquest.com/docview/2435008176?pq-origsite=%requestingapplication% |
PQID | 2435008176 |
PQPubID | 2050157 |
ParticipantIDs | proquest_journals_2435008176 |
PublicationCentury | 2000 |
PublicationDate | 20200817 |
PublicationDateYYYYMMDD | 2020-08-17 |
PublicationDate_xml | – month: 08 year: 2020 text: 20200817 day: 17 |
PublicationDecade | 2020 |
PublicationPlace | Ithaca |
PublicationPlace_xml | – name: Ithaca |
PublicationTitle | arXiv.org |
PublicationYear | 2020 |
Publisher | Cornell University Library, arXiv.org |
Publisher_xml | – name: Cornell University Library, arXiv.org |
SSID | ssj0002672553 |
Score | 3.2881815 |
SecondaryResourceType | preprint |
Snippet | Despite deep neural network (DNN)'s impressive prediction performance in various domains, it is well known now that a set of DNN models trained with the same... |
SourceID | proquest |
SourceType | Aggregation Database |
SubjectTerms | Activation Artificial neural networks Benchmarks Buckets Machine learning Randomness Recommender systems Training |
Title | Beyond Point Estimate: Inferring Ensemble Prediction Variation from Neuron Activation Strength in Recommender Systems |
URI | https://www.proquest.com/docview/2435008176 |
hasFullText | 1 |
inHoldings | 1 |
isFullTextHit | |
isPrint | |
link | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwfV3dS8MwED90RfDNT_yY40Bfh1uTNasvotI6hY0iKnsbXZrqwKaz7V79283FVh-EPYZASI7k7n53ufsBXDDjdSs_JphKJTmUa_TT2IAVPyaKGpdx2-x5PPFGL_xxOpjWAbey_lbZ6ESrqJNcUoz80jV2neyX8K6Xn11ijaLsak2hsQlO3xWCbvUwvP-NsbieMB4z-6dmre0Id8CJ4qUqdmFD6T3Ysl8uZbkPq5_iEYzyha4wME_NOI_qCh-oBI-ibRjoUmXzD4VRQekUEiG-GmxrhYlUGIK2t4bGG9mwlCGlmfVb9Y4LjYQts8ySxWHdmvwAzsPg-W7UbfY6q29TOfs7OzuEls61OgIcsAGTPcmlUAnn8dCXMaG9tC_SxO9xdQztdSudrJ8-hW2XkCU1fxVtaFXFSp0Z81vNO1bGHXBug0n0ZEbjr-Ab0kiQ6w |
link.rule.ids | 783,787,12779,21402,33387,33758,43614,43819 |
linkProvider | ProQuest |
linkToHtml | http://utb.summon.serialssolutions.com/2.0.0/link/0/eLvHCXMwfV1LS8NAEB60QfTmEx9VB_RarNk8jBdRSUm1DUGq9BbSzUYLZlOT9P-7syZ6EHoOLNlhd2a-mZ3vA7hkKusWXkIwlUZyqNfoZYkCK15CEjUmszTZ8zh0glfraWpPm4Jb1TyrbH2idtRpwalGfmWquE7xy3XuFl89Uo2i7mojobEOBlFVqVNtPPhh9PJbZTEdV-XM7J-j1dFjsA1GlCxEuQNrQu7Chn50yas9WP6Mj2BUzGWNvrpsKn0UtzikITyqt6EvK5HPPgVGJTVUyIj4ptCtNifSaAhqdg2J97zVKUNqNMv3-gPnEgld5rmWi8OGnHwfLgb-5DHotf8aN-epiv92zw6gIwspDgFtZjPe5xZ3RWpZyY3HE8J72bWbpV7fEkfQXbXS8erP57AZTMajeDQMn09gyyScSVSwbhc6dbkUpyoY17OzxuLfNDqScQ |
openUrl | ctx_ver=Z39.88-2004&ctx_enc=info%3Aofi%2Fenc%3AUTF-8&rfr_id=info%3Asid%2Fsummon.serialssolutions.com&rft_val_fmt=info%3Aofi%2Ffmt%3Akev%3Amtx%3Ajournal&rft.genre=article&rft.atitle=Beyond+Point+Estimate%3A+Inferring+Ensemble+Prediction+Variation+from+Neuron+Activation+Strength+in+Recommender+Systems&rft.jtitle=arXiv.org&rft.au=Chen%2C+Zhe&rft.au=Wang%2C+Yuyan&rft.au=Lin%2C+Dong&rft.au=Derek+Zhiyuan+Cheng&rft.date=2020-08-17&rft.pub=Cornell+University+Library%2C+arXiv.org&rft.eissn=2331-8422 |