4 Search Results for "Goldszmidt, Moises"

Document
DOI: 10.4230/OASIcs.NINeS.2026.22
There Is More to Internet Invariants Than Meets the Eye

Authors: Chris Misa, Walter Willinger, Ramakrishnan Durairajan, and Reza Rejaie

Published in: OASIcs, Volume 139, 1st New Ideas in Networked Systems (NINeS 2026)

Abstract
A rich body of literature assembled over the last 30 years shows that traffic traversing wide-area Internet links is consistent with self-similar (temporal) scaling behavior and that sets of observed addresses have multifractal (spatial) scaling behavior. In view of this empirical evidence, these behaviors cannot be viewed as mere mathematical curiosities but should justifiably be called invariants of measured Internet traffic (Internet invariants, for short). At the same time, it is fair to say that the early architects of the Internet were largely unaware of these properties and certainly did not intend to design a network so that the traffic traversing its links would exhibit self-similar scaling in time or multifractal scaling in the IP address space. In this paper, we resolve this apparent disconnect between architectural intentions and observed behaviors by applying a three-part framework that leverages, at its core, the perspective of Highly Optimized Tolerance (HOT). In particular, we take inspiration from studies on the origins of (temporal) self-similarity in measured Internet traffic but focus on a fundamentally new approach to understanding multifractal (spatial) scaling behavior. Specifically, we examine whether this invariant can be viewed as a visible hallmark of underlying but largely unknown robust design efforts, and explore a reverse-engineering approach to determine the concrete nature of the constrained optimization problems that these robust designs solve. Based on the insights gained from such reverse-engineering efforts, we speculate on the benefits of future efforts at forward-engineering - systematically leveraging the identified robust designs in order to provide scientifically sound intellectual foundations and practical principles for designing future networked systems.
Cite as

Chris Misa, Walter Willinger, Ramakrishnan Durairajan, and Reza Rejaie. There Is More to Internet Invariants Than Meets the Eye. In 1st New Ideas in Networked Systems (NINeS 2026). Open Access Series in Informatics (OASIcs), Volume 139, pp. 22:1-22:26, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2026)

Copy BibTex To Clipboard

@InProceedings{misa_et_al:OASIcs.NINeS.2026.22,
  author =	{Misa, Chris and Willinger, Walter and Durairajan, Ramakrishnan and Rejaie, Reza},
  title =	{{There Is More to Internet Invariants Than Meets the Eye}},
  booktitle =	{1st New Ideas in Networked Systems (NINeS 2026)},
  pages =	{22:1--22:26},
  series =	{Open Access Series in Informatics (OASIcs)},
  ISBN =	{978-3-95977-414-7},
  ISSN =	{2190-6807},
  year =	{2026},
  volume =	{139},
  editor =	{Argyraki, Katerina and Panda, Aurojit},
  publisher =	{Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{https://drops.dagstuhl.de/entities/document/10.4230/OASIcs.NINeS.2026.22},
  URN =		{urn:nbn:de:0030-drops-256077},
  doi =		{10.4230/OASIcs.NINeS.2026.22},
  annote =	{Keywords: Internet traffic, self-similarity, multifractal scaling, reverse-engineering}
}
Document
DOI: 10.4230/LIPIcs.FORC.2025.22
When Does a Predictor Know Its Own Loss?

Authors: Aravind Gollakota, Parikshit Gopalan, Aayush Karan, Charlotte Peale, and Udi Wieder

Published in: LIPIcs, Volume 329, 6th Symposium on Foundations of Responsible Computing (FORC 2025)

Abstract
Given a predictor and a loss function, how well can we predict the loss that the predictor will incur on an input? This is the problem of loss prediction, a key computational task associated with uncertainty estimation for a predictor. In a classification setting, a predictor will typically predict a distribution over labels and hence have its own estimate of the loss that it will incur, given by the entropy of the predicted distribution. Should we trust this estimate? In other words, when does the predictor know what it knows and what it does not know? In this work we study the theoretical foundations of loss prediction. Our main contribution is to establish tight connections between nontrivial loss prediction and certain forms of multicalibration [Ursula Hébert-Johnson et al., 2018], a multigroup fairness notion that asks for calibrated predictions across computationally identifiable subgroups. Formally, we show that a loss predictor that is able to improve on the self-estimate of a predictor yields a witness to a failure of multicalibration, and vice versa. This has the implication that nontrivial loss prediction is in effect no easier or harder than auditing for multicalibration. We support our theoretical results with experiments that show a robust positive correlation between the multicalibration error of a predictor and the efficacy of training a loss predictor.
Cite as

Aravind Gollakota, Parikshit Gopalan, Aayush Karan, Charlotte Peale, and Udi Wieder. When Does a Predictor Know Its Own Loss?. In 6th Symposium on Foundations of Responsible Computing (FORC 2025). Leibniz International Proceedings in Informatics (LIPIcs), Volume 329, pp. 22:1-22:22, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2025)

Copy BibTex To Clipboard

@InProceedings{gollakota_et_al:LIPIcs.FORC.2025.22,
  author =	{Gollakota, Aravind and Gopalan, Parikshit and Karan, Aayush and Peale, Charlotte and Wieder, Udi},
  title =	{{When Does a Predictor Know Its Own Loss?}},
  booktitle =	{6th Symposium on Foundations of Responsible Computing (FORC 2025)},
  pages =	{22:1--22:22},
  series =	{Leibniz International Proceedings in Informatics (LIPIcs)},
  ISBN =	{978-3-95977-367-6},
  ISSN =	{1868-8969},
  year =	{2025},
  volume =	{329},
  editor =	{Bun, Mark},
  publisher =	{Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{https://drops.dagstuhl.de/entities/document/10.4230/LIPIcs.FORC.2025.22},
  URN =		{urn:nbn:de:0030-drops-231490},
  doi =		{10.4230/LIPIcs.FORC.2025.22},
  annote =	{Keywords: loss prediction, multicalibration, active learning, algorithmic fairness, calibration, predictive uncertainty, uncertainty estimation, machine learning theory}
}
Document
DOI: 10.4230/LITES-v005-i001-a001
Risk-Aware Scheduling of Dual Criticality Job Systems Using Demand Distributions

Authors: Bader Naim Alahmad and Sathish Gopalakrishnan

Published in: LITES, Volume 5, Issue 1 (2018). Leibniz Transactions on Embedded Systems, Volume 5, Issue 1

Abstract
We pose the problem of scheduling Mixed Criticality (MC) job systems when there are only two criticality levels, Lo and Hi -referred to as Dual Criticality job systems- on a single processing platform, when job demands are probabilistic and their distributions are known. The current MC models require that the scheduling policy allocate as little execution time as possible to Lo-criticality jobs if the scenario of execution is of Hi criticality, and drop Lo-criticality jobs entirely as soon as the execution scenario's criticality level can be inferred and is Hi. The work incurred by "incorrectly" scheduling Lo-criticality jobs in cases of Hi realized scenarios might affect the feasibility of Hi criticality jobs; we quantify this work and call it Work Threatening Feasibility (WTF). Our objective is to construct online scheduling policies that minimize the expected WTF for the given instance, and under which the instance is feasible in a probabilistic sense that is consistent with the traditional deterministic definition of MC feasibility. We develop a probabilistic framework for MC scheduling, where feasibility is defined in terms of (chance) constraints on the probabilities that Lo and Hi jobs meet their deadlines. The probabilities are computed over the set of sample paths, or trajectories, induced by executing the policy, and those paths are dependent upon the set of execution scenarios and the given demand distributions. Our goal is to exploit the information provided by job distributions to compute the minimum expected WTF below which the given instance is not feasible in probability, and to compute a (randomized) "efficiently implementable" scheduling policy that realizes the latter quantity. We model the problem as a Constrained Markov Decision Process (CMDP) over a suitable state space and a finite planning horizon, and show that an optimal (non-stationary) Markov randomized scheduling policy exists. We derive an optimal policy by solving a Linear Program (LP). We also carry out quantitative evaluations on select probabilistic MC instances to demonstrate that our approach potentially outperforms current MC scheduling policies.
Cite as

Bader Naim Alahmad and Sathish Gopalakrishnan. Risk-Aware Scheduling of Dual Criticality Job Systems Using Demand Distributions. In LITES, Volume 5, Issue 1 (2018). Leibniz Transactions on Embedded Systems, Volume 5, Issue 1, pp. 01:1-01:30, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2018)

Copy BibTex To Clipboard

@Article{alahmad_et_al:LITES-v005-i001-a001,
  author =	{Alahmad, Bader Naim and Gopalakrishnan, Sathish},
  title =	{{Risk-Aware Scheduling of Dual Criticality Job Systems Using Demand Distributions}},
  journal =	{Leibniz Transactions on Embedded Systems},
  pages =	{01:1--01:30},
  ISSN =	{2199-2002},
  year =	{2018},
  volume =	{5},
  number =	{1},
  publisher =	{Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{https://drops.dagstuhl.de/entities/document/10.4230/LITES-v005-i001-a001},
  URN =		{urn:nbn:de:0030-drops-192720},
  doi =		{10.4230/LITES-v005-i001-a001},
  annote =	{Keywords: Mixed criticalities, Probability distribution, Real time systems, Scheduling, Chance constrained Markov decision process, Linear programming, Randomized policy}
}
Document
DOI: 10.4230/DagSemProc.09201.8
Wheels within Wheels: Making Fault Management Cost-Effective

Authors: Moises Goldszmidt, Miroslaw Malek, Simin Nadjm-Tehrani, Priya Narasimhan, Felix Salfner, Paul A.S. Ward, and John Wilkes

Published in: Dagstuhl Seminar Proceedings, Volume 9201, Self-Healing and Self-Adaptive Systems (2009)

Abstract
Local design and optimization of the components of a fault management system results in sub-optimal decisions. This means that the target system will likely not meet its objectives (under-performs) or cost too much if conditions, objectives, or constraints change. We can fix this by applying a nested, management system for the fault-management system itself. We believe that doing so will produce a more resilient, self-aware, system that can operate more effectively across a wider range of conditions, and provide better behavior at closer to optimal cost. This document summarizes the results of the Working Group 7 - ``Cost-Effective Fault Management'' - at the Dagstuhl Seminar 09201 ``Self-Healing and Self-Adaptive Systems'' (organized by A. Andrzejak, K. Geihs, O. Shehory and J. Wilkes). The seminar was held from May 10th 2009 to May 15th 2009 in Schloss Dagstuhl~--~Leibniz Center for Informatics.
Cite as

Moises Goldszmidt, Miroslaw Malek, Simin Nadjm-Tehrani, Priya Narasimhan, Felix Salfner, Paul A.S. Ward, and John Wilkes. Wheels within Wheels: Making Fault Management Cost-Effective. In Self-Healing and Self-Adaptive Systems. Dagstuhl Seminar Proceedings, Volume 9201, pp. 1-6, Schloss Dagstuhl – Leibniz-Zentrum für Informatik (2009)

Copy BibTex To Clipboard

@InProceedings{goldszmidt_et_al:DagSemProc.09201.8,
  author =	{Goldszmidt, Moises and Malek, Miroslaw and Nadjm-Tehrani, Simin and Narasimhan, Priya and Salfner, Felix and Ward, Paul A.S. and Wilkes, John},
  title =	{{Wheels within Wheels: Making Fault Management Cost-Effective}},
  booktitle =	{Self-Healing and Self-Adaptive Systems},
  pages =	{1--6},
  series =	{Dagstuhl Seminar Proceedings (DagSemProc)},
  ISSN =	{1862-4405},
  year =	{2009},
  volume =	{9201},
  editor =	{Artur Andrzejak and Kurt Geihs and Onn Shehory and John Wilkes},
  publisher =	{Schloss Dagstuhl -- Leibniz-Zentrum f{\"u}r Informatik},
  address =	{Dagstuhl, Germany},
  URL =		{https://drops.dagstuhl.de/entities/document/10.4230/DagSemProc.09201.8},
  URN =		{urn:nbn:de:0030-drops-21029},
  doi =		{10.4230/DagSemProc.09201.8},
  annote =	{Keywords: Fault management, cost-effectiveness}
}
  • Refine by Type
  • 4 Document/PDF
  • 2 Document/HTML

  • Refine by Publication Year
  • 1 2026
  • 1 2025
  • 1 2018
  • 1 2009

  • Refine by Author
  • 1 Alahmad, Bader Naim
  • 1 Durairajan, Ramakrishnan
  • 1 Goldszmidt, Moises
  • 1 Gollakota, Aravind
  • 1 Gopalakrishnan, Sathish
  • Show More...

  • Refine by Series/Journal
  • 1 LIPIcs
  • 1 OASIcs
  • 1 LITES
  • 1 DagSemProc

  • Refine by Classification
  • 1 Mathematics of computing → Markov processes
  • 1 Mathematics of computing → Probability and statistics
  • 1 Networks → Network structure
  • 1 Networks → Public Internet
  • 1 Software and its engineering → Real-time schedulability
  • Show More...

  • Refine by Keyword
  • 1 Chance constrained Markov decision process
  • 1 Fault management
  • 1 Internet traffic
  • 1 Linear programming
  • 1 Mixed criticalities
  • Show More...

Any Issues?
X

Feedback on the Current Page

CAPTCHA

Thanks for your feedback!

Feedback submitted to Dagstuhl Publishing

Could not send message

Please try again later or send an E-mail
© 2023-2026 Schloss Dagstuhl – LZI GmbH Schloss Dagstuhl – LZI GmbH About DROPS Imprint Privacy Contact