A Refined Mean Field Approximation

@article{Gast2017ARM,
  title={A Refined Mean Field Approximation},
  author={Nicolas Gast and Benny Van Houdt},
  journal={Proceedings of the ACM on Measurement and Analysis of Computing Systems},
  year={2017},
  volume={1},
  pages={1 - 28}
}
  • Nicolas Gast, B. V. Houdt
  • Published 19 December 2017
  • Computer Science
  • Proceedings of the ACM on Measurement and Analysis of Computing Systems
Mean field models are a popular means to approximate large and complex stochastic models that can be represented as N interacting objects. Recently it was shown that under very general conditions the steady-state expectation of any performance functional converges at rate O(1/N) to its mean field approximation. In this paper we establish a result that expresses the constant associated with this 1/N term. This constant can be computed easily as it is expressed in terms of the Jacobian and… 

Figures and Tables from this paper

Size Expansions of Mean Field Approximation: Transient and Steady-State Analysis
TLDR
This paper shows how to obtain the same result for the transient regime and provides a further refinement by expanding the term in 1/N2 (both for transient and steady-state regime) and provides examples that show this new approximation is usable in practice for systems with up to a few tens of dimensions.
Refining Mean-field Approximations by Dynamic State Truncation
TLDR
A new method is presented that couples the master equation governing the evolution of the probability distribution of a truncation of the original state space with a mean-field approximation of a time-inhomogeneous population process that dynamically shifts the truncation across the whole state space and provides a result of asymptotic correctness in the limit when the truncations covers the state space.
Refining Mean-field Approximations by Dynamic State Truncation
TLDR
A new method is presented that couples the master equation governing the evolution of the probability distribution of a truncation of the original state space with a mean-field approximation of a time-inhomogeneous population process that dynamically shifts the truncation across the whole state space and provides a result of asymptotic correctness in the limit when the truncations covers the state space.
Size expansions of mean field approximation: Transient and steady-state analysis
A refined mean field approximation of synchronous discrete-time population models
A Refined Mean Field Approximation for Synchronous Population Processes
TLDR
It is shown that, similarly to the asynchronous case, the mean field approximation of a synchronous population can be refined by a term in 1/N, and holds for finite time-horizon and steady-state.
Mean Field and Refined Mean Field Approximations for Heterogeneous Systems
TLDR
An interaction model is defined that allows obtaining asymptotic convergence results for stochastic systems with heterogeneous object behavior, and it is shown that for moderate system sizes the refined mean field approximation tends to be more accurate than simulations for any reasonable simulation time.
A Refined Mean Field Approximation
TLDR
This paper focuses on mean field theory, a very popular approximation method used to study the behavior of stochastic models used to assess the performance of computer systems for many decades.
A Refined Mean Field Approximation
  • Nicolas Gast, B. Van Houdt
  • Computer Science
    Abstracts of the 2018 ACM International Conference on Measurement and Modeling of Computer Systems
  • 2018
TLDR
This paper focuses on mean field theory, a very popular approximation method used to study the behavior of stochastic models used to assess the performance of computer (and other) systems for many decades.
A Refined Mean Field Approximation
TLDR
This paper focuses on mean field theory, a very popular approximation method used to study the behavior of stochastic models used to assess the performance of computer systems for many decades.
...
1
2
3
4
5
...

References

SHOWING 1-10 OF 38 REFERENCES
Expected Values Estimated via Mean-Field Approximation are 1/N-Accurate
  • Nicolas Gast
  • Computer Science, Mathematics
    Proc. ACM Meas. Anal. Comput. Syst.
  • 2017
TLDR
This paper revisits the accuracy of mean-field approximation by focusing on expected values and develops a new perturbation theory argument that shows that the result holds for the stationary regime if the dynamical system is asymptotically exponentially stable.
On the Approximation Error of Mean-Field Models
  • Lei Ying
  • Computer Science
    SIGMETRICS 2016
  • 2016
TLDR
A general theorem is established that under some mild conditions, if the mean-field model is globally asymptotically stable and locally exponentially stable, the mean square difference between the stationary distribution of the stochastic system with size M and the equilibrium point of the corresponding mean- field system is O(1/M).
A Refined Mean Field Approximation
TLDR
This paper focuses on mean field theory, a very popular approximation method used to study the behavior of stochastic models used to assess the performance of computer systems for many decades.
Stein's Method for Mean-Field Approximations in Light and Heavy Traffic Regimes
  • Lei Ying
  • Computer Science
    SIGMETRICS 2017
  • 2017
TLDR
This paper takes a different point of view on mean-field models, regarding a mean- field model as the limiting system of large-scale stochastic system, and views the equilibrium point of themean-field model, called a mean -field solution, simply as an approximation of the stationary distribution of the finite-size system.
A Generic Mean Field Convergence Result for Systems of Interacting Objects
We consider a model for interacting objects, where the evolution of each object is given by a finite state Markov chain, whose transition matrix depends on the present and the past of the
Strong approximation theorems for density dependent Markov chains
Diffusion models and steady-state approximations for exponentially ergodic Markovian queues.
Motivated by queues with many servers, we study Brownian steady-state approximations for continuous time Markov chains (CTMCs). Our approximations are based on diffusion models (rather than a
A class of mean field interaction models for computer and communication systems
  • M. Benaim, J. Le Boudec
  • Mathematics
    2008 6th International Symposium on Modeling and Optimization in Mobile, Ad Hoc, and Wireless Networks and Workshops
  • 2008
Markov chains with discontinuous drifts have differential inclusion limits
Stein's method for steady-state diffusion approximations of $M/Ph/n+M$ systems
TLDR
It is proved that the Wasserstein distance between the stationary distribution of the normalized system size process and that of a piecewise Ornstein-Uhlenbeck (OU) process is bounded by $C/\sqrt{\lambda}$, where the constant $C$ is independent of the arrival rate $\lambda$ and the number of servers as long as they are in the Halfin-Whitt parameter regime.
...
1
2
3
4
...