Filters








5,112 Hits in 5.2 sec

Unbiased Deterministic Total Ordering of Parallel Simulations with Simultaneous Events [article]

Neil McGlohon, Christopher D. Carothers
2021 arXiv   pre-print
This is synonymous with finding a consistent total ordering of events.  ...  We implement and evaluate this extended definition of virtual time in the Rensselaer Optimistic Simulation System (ROSS) with three simulation models and discuss the importance of deterministic event ordering  ...  In this work we propose and argue three new mechanisms for finding a deterministic total ordering of events in a parallel discrete event simulation with simultaneous events: (1) Unbiased random total ordering  ... 
arXiv:2105.00069v1 fatcat:l5bx3td66bhajmqucvmcgnop2y

Importance Sampling for Polarization-Mode Dispersion: Techniques and Applications

G. Biondini, W.L. Kath, C.R. Menyuk
2004 Journal of Lightwave Technology  
The use of IS allows rare PMD events to be simulated much more efficiently than with standard MC methods.  ...  The basic theory of importance sampling (IS) as relevant to polarization-mode dispersion (PMD) in optical fibers is discussed, and its application to Monte Carlo (MC) simulations of PMD-induced transmission  ...  The technique allows low-probability events to be efficiently simulated by concentrating MC simulations in the most significant regions of interest in sample space.  ... 
doi:10.1109/jlt.2004.825321 fatcat:zarrrss2sjb3dbauyuvcxns4d4

Efficient Simultaneous Simulation of Markov Chains [chapter]

Carsten Wächter, Alexander Keller
2008 Monte Carlo and Quasi-Monte Carlo Methods 2006  
Then we discuss the efficiency of different sorting strategies for the example of light transport simulation.  ...  Given an order on the state space, both approaches can be improved by sorting the ensemble of Markov chains.  ...  Acknowledgments The authors would like to thank mental images GmbH for support and for funding of this research. Jan Rzehak created the illustration in Figure 3 .  ... 
doi:10.1007/978-3-540-74496-2_41 fatcat:ggc7uyrajnawff6wadcqwkxadi

Past–future information bottleneck framework for simultaneously sampling biomolecular reaction coordinate, thermodynamics and kinetics [article]

Yihang Wang, João Marcelo Lamim Ribeiro, Pratyush Tiwary
2018 bioRxiv   pre-print
, especially when plagued with rare events.  ...  In this work we draw parallels between such tasks, and the efficient sampling of complex biomolecules with hundreds of thousands of atoms.  ...  PT would like to thank the University of Maryland Graduate School for financial support through the Research and Scholarship Award (RASA).  ... 
doi:10.1101/507822 fatcat:pg5o2gwugjeefexwd6ruhrbmu4

Stochastic optimization and the simultaneous perturbation method

James C. Spall
1999 Proceedings of the 31st conference on Winter simulation Simulation---a bridge to the future - WSC '99  
This feature allows for a significant decrease in the cost of optimization, especially in problems with a large number of variables to be optimized.  ...  In the above spirit, the "simultaneous perturbation stochastic approximation (SPSA)" method for difficult multivariate optimization problems has been developed.  ...  The second-order ASP approach is composed of two parallel recursions: one for θ and one for the Hessian of L(θ).  ... 
doi:10.1145/324138.324170 fatcat:55i2jup7n5cw5euahlqi2gpecy

Asynchronous Anytime Sequential Monte Carlo [article]

Brooks Paige, Frank Wood, Arnaud Doucet, Yee Whye Teh
2014 arXiv   pre-print
It is an anytime algorithm in the sense that it can be run forever to emit an unbounded number of particles while keeping within a fixed memory budget.  ...  We prove that the particle cascade is an unbiased marginal likelihood estimator which means that it can be straightforwardly plugged into existing pseudomarginal methods.  ...  Air Force Research Laboratory of the U.S. Government.  ... 
arXiv:1407.2864v1 fatcat:ux7lcccoprd2pbrglpld2bsmtu

Multiple importance sampling for first- and second-order polarization-mode dispersion

S.L. Fogal, G. Biondini, W.L. Kath
2002 IEEE Photonics Technology Letters  
A simulation method that targets all possible combinations of first-and second-order polarization-mode dispersion (PMD) is described.  ...  Use of this method in importance-sampled Monte Carlo simulations yields a more comprehensive determination of PMD-induced system penalties than first-order biasing alone and significantly speeds up the  ...  The dashed line shows the much smaller region obtained with unbiased samples. Of course, a more complete coverage of the -plane is often needed.  ... 
doi:10.1109/lpt.2002.801091 fatcat:ikyck523uremdcrmadfu4jib34

PASTO: Strategic Parameter Optimization in Recommendation Systems – Probabilistic is Better than Deterministic [article]

Weicong Ding, Hanlin Tang, Jingshuo Feng, Lei Yuan, Sen Yang, Guangxu Yang, Jie Zheng, Jing Wang, Qiang Su, Dong Zheng, Xuezhong Qiu, Yongqi Liu (+8 others)
2021 arXiv   pre-print
In this setting with multiple and constrained goals, this paper discovers that a probabilistic strategic parameter regime can achieve better value compared to the standard regime of finding a single deterministic  ...  Our approach is applied in a popular social network platform with hundreds of millions of daily users and achieves +0.22% lift of user engagement in a recommendation task and +1.7% lift in revenue in an  ...  The proposed approach, Probabilistic pArameter optimization with unbiased STOchastic gradient approximation (PASTO), converges at a rate O 1/ √ T and admits the regret in an order of O √ T .  ... 
arXiv:2108.09076v1 fatcat:wnhgbegpabaihh6rlesgrlxq4e

An excited-state approach within full configuration interaction quantum Monte Carlo

N. S. Blunt, Simon D. Smart, George H. Booth, Ali Alavi
2015 Journal of Chemical Physics  
We present a new approach to calculate excited states with the full configuration interaction quantum Monte Carlo (FCIQMC) method.  ...  It can thus be used to study several of the lowest-energy states of a system within the same symmetry.  ...  Simulations used 1.6 × 10 7 walkers per state, with a deterministic space of size D = 10 4 and a trial space of size T = 2 × 10 3 .  ... 
doi:10.1063/1.4932595 pmid:26450302 fatcat:qua27fr6c5fsppd4lzlskrx7va

Past–future information bottleneck for sampling molecular reaction coordinate simultaneously with thermodynamics and kinetics

Yihang Wang, João Marcelo Lamim Ribeiro, Pratyush Tiwary
2019 Nature Communications  
In this work we draw parallels between these and the efficient sampling of biomolecules with hundreds of thousands of atoms.  ...  For this we use the Predictive Information Bottleneck framework used for the first two problems, and re-formulate it for the sampling of biomolecules, especially when plagued with rare events.  ...  P.T. would like to thank the University of Maryland Graduate School for financial support through the Research and Scholarship Award (RASA).  ... 
doi:10.1038/s41467-019-11405-4 pmid:31395868 pmcid:PMC6687748 fatcat:vb3r3qzj2neufl3mgfifp3eop4

Reachability Analysis for Stochastic Hybrid Systems Using Multilevel Splitting [chapter]

Derek Riley, Xenofon Koutsoukos, Kasandra Riley
2009 Lecture Notes in Computer Science  
In Table 1 we compare the variance and execution times using an order 0.5 simulation method (sim1) and an order 1.0 method with probabilistic boundary detection (sim2) from [4] with both MC methods  ...  trajectories, and the total number of trajectories n m is incremented by j − 1.  ... 
doi:10.1007/978-3-642-00602-9_35 fatcat:kgouu4mtcvgmjekfx4wlvtjhs4

Optimization via simulation: A review

Michael C. Fu
1994 Annals of Operations Research  
We review techniques for optimizing stochastic discrete-event systems via simulation.  ...  is the upper α quantile of a random variable that is the maximum of K − 1 equally correlated multivariate-t random variables with correlation 0.5 and K(n − 1) degrees of freedom (found in tables of Hochberg  ...  It is hoped that the recent surge of interest, coupled with the advent of parallel simulation, will soon change this situation.  ... 
doi:10.1007/bf02136830 fatcat:srcum3kqvzh33gbd4g3kghncje

Evaluating Performance of Spintronics-Based Spiking Neural Network Chips using Parallel Discrete Event Simulation

Elkin Cruz-Camacho, Siyuan Qian, Ankit Shukla, Neil McGlohon, Shaloo Rakheja, Christopher D. Carothers
2022 SIGSIM Conference on Principles of Advanced Discrete Simulation  
To explore the benefits of spintronics-based hardware on realistic neuromorphic workloads, we developed a Parallel Discrete-Event Simulation model called Doryta, which is further integrated with a materials-to-systems  ...  Through Doryta we are also able to show the power of neuromorphic computing by simulating Conway's Game of Life.  ...  The views and conclusions contained herein are those of the authors and should not be interpreted as necessarily representing the official policies or endorsements, either expressed or implied, of Air  ... 
doi:10.1145/3518997.3531025 fatcat:7da5e3nwlncw5gbdejimaceiiy

Unknown mixing times in apprenticeship and reinforcement learning [article]

Tom Zahavy, Alon Cohen, Haim Kaplan, Yishay Mansour
2020 arXiv   pre-print
To illustrate the difficulty, consider a periodic Markov chain with states ordered in a (deterministic) cycle.  ...  Intuitively, CFTP starts |S| parallel simulations of the Markov chain, one from each state, at minus infinity. When two simulations reach the same state, they continue together as one simulation.  ...  On this event, we continue simulating the chains until the end of the next block and continue in this manner. This is analogous to the following balls-and-bins process.  ... 
arXiv:1905.09704v2 fatcat:7hfh5ygulzcrfjzocpzh3q464e

Feature Article: Optimization for simulation: Theory vs. Practice

Michael C. Fu
2002 INFORMS journal on computing  
P robably one of the most successful interfaces between operations research and computer science has been the development of discrete-event simulation software.  ...  The main thesis of this article, however, is that there is a disconnect between research in simulation optimization-which has addressed the stochastic nature of discrete-event simulation by concentrating  ...  Acknowledgments This work was supported in part by the National Science Foundation under Grants DMI-9713720 and DMI-9988867 and by the Air Force Office of Scientific Research under Grant F496200110161.  ... 
doi:10.1287/ijoc.14.3.192.113 fatcat:ntus2mrdhjbptc77jf2zjmgmcu
« Previous Showing results 1 — 15 out of 5,112 results