Powered by OpenAIRE graph
Found an issue? Give us feedback
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/ https://zenodo.org/r...arrow_drop_down
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
https://zenodo.org/record/1159...
Article
License: CC BY
Data sources: UnpayWall
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
ZENODO
Conference object . 2018
License: CC BY
Data sources: Datacite
image/svg+xml art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos Open Access logo, converted into svg, designed by PLoS. This version with transparent background. http://commons.wikimedia.org/wiki/File:Open_Access_logo_PLoS_white.svg art designer at PLoS, modified by Wikipedia users Nina, Beao, JakobVoss, and AnonMoos http://www.plos.org/
ZENODO
Conference object . 2018
License: CC BY
Data sources: Datacite
https://doi.org/10.23919/eusip...
Article . 2017 . Peer-reviewed
Data sources: Crossref
https://dx.doi.org/10.60692/ek...
Other literature type . 2017
Data sources: Datacite
https://dx.doi.org/10.60692/vh...
Other literature type . 2017
Data sources: Datacite
versions View all 5 versions
addClaim

This Research product is the result of merged Research products in OpenAIRE.

You have already added 0 works in your ORCID record related to the merged Research product.

Online convex optimization for dynamic network resource allocation

تحسين محدب عبر الإنترنت لتخصيص موارد الشبكة الديناميكية
Authors: Tianyi Chen; Qing Ling; Georgios B. Giannakis;

Online convex optimization for dynamic network resource allocation

Abstract

Le présent article traite de l'optimisation convexe en ligne impliquant des fonctions de perte contradictoire et des contraintes contradictoires, où les contraintes sont révélées après avoir pris des décisions, et peuvent être tolérables à des violations instantanées mais doivent être satisfaites à long terme. La performance d'un algorithme en ligne dans ce contexte est évaluée par : i) la différence de ses pertes par rapport à la meilleure solution dynamique avec des informations à un créneau de la fonction de perte et de la contrainte (qui est ici appelée regret dynamique) ; et ii) le nombre accumulé de violations de contraintes (qui est ici appelé ajustement dynamique). Dans ce contexte, un schéma de point de selle en ligne modifié (MOSP) est développé et s'est avéré produire simultanément un regret et un ajustement dynamiques sous-linéaires, à condition que les variations accumulées des minimiseurs et des contraintes par créneau augmentent sous-linéairement avec le temps. Le MOSP est appliqué à la tâche d'allocation dynamique des ressources du réseau, et il est démontré qu'il surpasse la méthode bien connue du double gradient stochastique.

El presente documento trata sobre la optimización convexa en línea que implica funciones de pérdida adversaria y restricciones adversarias, donde las restricciones se revelan después de tomar decisiones, y pueden ser tolerables a violaciones instantáneas, pero deben satisfacerse a largo plazo. El rendimiento de un algoritmo en línea en este entorno se evalúa por: i) la diferencia de sus pérdidas en relación con la mejor solución dinámica con información de una ranura de la función de pérdida y la restricción (que aquí se denomina arrepentimiento dinámico); y, ii) la cantidad acumulada de violaciones de restricciones (que aquí se denomina ajuste dinámico). En este contexto, se desarrolla un esquema modificado de punto de silla en línea (MOSP) y se demuestra que produce simultáneamente arrepentimiento y ajuste dinámico sublineal, siempre que las variaciones acumuladas de minimizadores y restricciones por ranura crezcan sublinealmente con el tiempo. El MOSP se aplica a la tarea de asignación dinámica de recursos de red y se muestra que supera al conocido método estocástico de doble gradiente.

The present paper deals with online convex optimization involving adversarial loss functions and adversarial constraints, where the constraints are revealed after making decisions, and can be tolerable to instantaneous violations but must be satisfied in the long term. Performance of an online algorithm in this setting is assessed by: i) the difference of its losses relative to the best dynamic solution with one-slot-ahead information of the loss function and the constraint (that is here termed dynamic regret); and, ii) the accumulated amount of constraint violations (that is here termed dynamic fit). In this context, a modified online saddle-point (MOSP) scheme is developed, and proved to simultaneously yield sub-linear dynamic regret and fit, provided that the accumulated variations of per-slot minimizers and constraints are sub-linearly growing with time. MOSP is applied to the dynamic network resource allocation task, and shown to outperform the well-known stochastic dual gradient method.

تتعامل هذه الورقة مع التحسين المحدب عبر الإنترنت الذي يتضمن وظائف الخسارة العدائية والقيود العدائية، حيث يتم الكشف عن القيود بعد اتخاذ القرارات، ويمكن تحملها للانتهاكات الفورية ولكن يجب تلبيتها على المدى الطويل. يتم تقييم أداء الخوارزمية عبر الإنترنت في هذا الإعداد من خلال: 1) اختلاف خسائرها بالنسبة إلى أفضل حل ديناميكي مع معلومات الشق الأمامي لوظيفة الخسارة والقيود (التي تسمى هنا الندم الديناميكي )؛ و 2) المبلغ المتراكم لانتهاكات القيود (التي تسمى هنا التوافق الديناميكي). في هذا السياق، تم تطوير مخطط نقطة السرج (MOSP) المعدل عبر الإنترنت، وأثبت أنه ينتج في وقت واحد ندمًا ديناميكيًا دون خطي وملاءمة، بشرط أن تكون الاختلافات المتراكمة في عوامل التقليل والقيود لكل فتحة تنمو خطيًا مع مرور الوقت. يتم تطبيق MOSP على مهمة تخصيص موارد الشبكة الديناميكية، ويظهر أنه يتفوق على طريقة التدرج المزدوج العشوائي المعروفة.

Keywords

Regret Analysis, Convex Optimization, Computer Networks and Communications, Economics, Social Sciences, Geometry, Evolutionary biology, Optimization of Multi-Armed Bandit Problems, Management Science and Operations Research, Decision Sciences, Systems engineering, Task (project management), Context (archaeology), Engineering, Convex function, Distributed Multi-Agent Coordination and Control, Artificial Intelligence, Distributed Optimization, Machine learning, Resource management (computing), Dynamic network analysis, FOS: Mathematics, Optimization Methods in Machine Learning, Constraint (computer-aided design), Optimization problem, Resource allocation, Biology, Economic growth, Computer network, Mathematical optimization, Paleontology, Bandit Optimization, Saddle point, Computer science, Distributed computing, Convex optimization, Regular polygon, Regret, Online Learning, Function (biology), Computer Science, Physical Sciences, Convergence (economics), Mathematics

  • BIP!
    Impact byBIP!
    selected citations
    These citations are derived from selected sources.
    This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    7
    popularity
    This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
    Average
    influence
    This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
    Average
    impulse
    This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
    Top 10%
    OpenAIRE UsageCounts
    Usage byUsageCounts
    visibility views 3
    download downloads 14
  • 3
    views
    14
    downloads
    Powered byOpenAIRE UsageCounts
Powered by OpenAIRE graph
Found an issue? Give us feedback
visibility
download
selected citations
These citations are derived from selected sources.
This is an alternative to the "Influence" indicator, which also reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Citations provided by BIP!
popularity
This indicator reflects the "current" impact/attention (the "hype") of an article in the research community at large, based on the underlying citation network.
BIP!Popularity provided by BIP!
influence
This indicator reflects the overall/total impact of an article in the research community at large, based on the underlying citation network (diachronically).
BIP!Influence provided by BIP!
impulse
This indicator reflects the initial momentum of an article directly after its publication, based on the underlying citation network.
BIP!Impulse provided by BIP!
views
OpenAIRE UsageCountsViews provided by UsageCounts
downloads
OpenAIRE UsageCountsDownloads provided by UsageCounts
7
Average
Average
Top 10%
3
14
Green