Filters








1,635 Hits in 7.9 sec

An Improved Generic Bet-and-Run Strategy for Speeding Up Stochastic Local Search [article]

Thomas Weise, Zijun Wu, Markus Wagner
2018 arXiv   pre-print
Building on the recent success of Bet-and-Run approaches for restarted local search solvers, we introduce an improved generic Bet-and-Run strategy.  ...  A commonly used strategy for improving optimization algorithms is to restart the algorithm when it is believed to be trapped in an inferior part of the search space.  ...  BET-AND-RUN Strategy for Speeding Up Stochastic Local Search 4 strategy up to constant factors.  ... 
arXiv:1806.08984v1 fatcat:cqayu46t5vg7rchata72b755sm

A Generic Bet-and-Run Strategy for Speeding Up Stochastic Local Search

Tobias Friedrich, Timo Kötzing, Markus Wagner
2017 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
However, while specific restart strategies have been developed for specific problems (and specific algorithms), restarts are typically not regarded as a general tool to speed up an optimization algorithm  ...  A common strategy for improving optimization algorithms is to restart the algorithm when it is believed to be trapped in an inferior part of the search space.  ...  Conclusions and Future Work We study a generic bet-and-run restart strategy, which is easy to implement as an additional speed-up heuristic for solving difficult optimization problems.  ... 
doi:10.1609/aaai.v31i1.10645 fatcat:w6tufyyi4bhpfhyy7wtqpkn7sq

Implementing the BBE Agent-Based Model of a Sports-Betting Exchange [article]

Dave Cliff and James Hawkins and James Keen and Roberto Lau-Soto
2021 arXiv   pre-print
develop and test new betting strategies via advanced data analytics and machine learning techniques.  ...  The motivation for constructing this ABM, which is known as the Bristol Betting Exchange (BBE), is so that it can serve as a synthetic data generator, producing large volumes of data that can be used to  ...  processors, for a /P speed-up.  ... 
arXiv:2108.02419v1 fatcat:3l62oo2p7vhrjc2rr53rpwuqea

An Improved Generic Bet-and-Run Strategy with Performance Prediction for Stochastic Local Search

Thomas Weise, Zijun Wu, Markus Wagner
2019 PROCEEDINGS OF THE THIRTIETH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE AND THE TWENTY-EIGHTH INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE  
Building on the recent success of BET-AND-RUN approaches for restarted local search solvers, we introduce a more generic version that makes use of performance prediction.  ...  A commonly used strategy for improving optimization algorithms is to restart the algorithm when it is believed to be trapped in an inferior part of the search space.  ...  Building on the success of BET-AND-RUN approaches for restarted local search solvers, (Kadioglu, Sellmann, and Wagner 2017) introduced the idea of adaptive restart strategies.  ... 
doi:10.1609/aaai.v33i01.33012395 fatcat:3r3cgmy7sfgnjaujj4qu4euxti

DIFFERENTIAL EVOLUTION APPROACH TO CALCULATE OPTIMAL RAMP METERING RATES

Anton Sysoev, Sandra Hohmann, Justin Geistefeldt
2017 International Journal for transport and traffic engineering  
The paper introduces DERMS -an approach for a coordinated ramp metering control strategy based on solving a non-linear optimization problem.  ...  The solution of the described problem was found using the Differential Evolution strategy giving a global optimum for non-linear and non-differentiable or multimodal functions.  ...  When various ramps are linked, they can be controlled connectedly to improve local strategies.  ... 
doi:10.7708/ijtte.2017.7(1).05 fatcat:wyftrtgtprhqpccu2g4ydswdjq

Search in Imperfect Information Games [article]

Martin Schmid
2021 arXiv   pre-print
Chess and AlphaGo for Go.  ...  TD-Gammon improves upon those ideas and uses neural networks to learn those complex value functions – only to be again used within search.  ...  While search reasons inherently locally, it needs to make sure the resulting strategy is strongly globally consistent with an optimal strategy for the full game.  ... 
arXiv:2111.05884v1 fatcat:lrwo6qdbmzgyri64625zd3vdjm

Better automated abstraction techniques for imperfect information games, with application to Texas Hold'em poker

Andrew Gilpin, Tuomas Sandholm
2007 Proceedings of the 6th international joint conference on Autonomous agents and multiagent systems - AAMAS '07  
The techniques lead to a drastic improvement over prior approaches for automatically generating agents, and our agent plays competitively even against the best agents overall.  ...  We present new approximation methods for computing gametheoretic strategies for sequential games of imperfect information. At a high level, we contribute two new ideas.  ...  Three community cards are dealt face-up, and a betting round takes place with bets equal to two chips.  ... 
doi:10.1145/1329125.1329358 dblp:conf/atal/GilpinS07 fatcat:ebuxhjl3z5gnddhxxf7rzwt4ne

Iterative Decomposition Guided Variable Neighborhood Search for Graphical Model Energy Minimization

Abdelkader Ouali, David Allouche, Simon de Givry, Samir Loudni, Yahia Lebbah, Lakhdar Loukil
2017 Conference on Uncertainty in Artificial Intelligence  
In this paper, we propose an iterative approach above VNS which uses (partial) tree search inside its local neighborhood exploration.  ...  Most complete methods rely on tree search, while incomplete methods rely on local search. Among them, we study Variable Neighborhood Search (VNS) for graphical models.  ...  We are grateful to the genotoul bioinformatics platform Toulouse Midi-Pyrenees (Bioinfo Genotoul) and the HPC HAYTHAM of university of Oran 1 for providing computing and storage resources.  ... 
dblp:conf/uai/OualiAGLLL17 fatcat:e5twqnknuff5ddeo6f6syoyhdu

Forecasting Price Movements in Betting Exchanges Using Cartesian Genetic Programming and ANN

Ivars Dzalbs, Tatiana Kalganova
2018 Big Data Research  
Although it implements some stochastic search heuristic in order to improve strategy performance, it does not use any machine learning techniques to develop the strategy itself.  ...  Data is then transferred to a local server on a weekly basis where it is formatted in open-source Protobuf-net for fast deserialization speeds and processing.  ... 
doi:10.1016/j.bdr.2018.10.001 fatcat:d64ncanx7jd65d7aq5bjxap7sa

Universal parameter optimisation in games based on SPSA

Levente Kocsis, Csaba Szepesvári
2006 Machine Learning  
The SPSA algorithm (Simultaneous Perturbation Stochastic Approximation) is a generic stochastic gradient method for optimising an objective function when an analytic expression of the gradient is not available  ...  As such, it is an attractive choice for parameter optimisation in game programs, both due to its generality and simplicity.  ...  Acknowledgments The authors wish to thank the reviewers for their many useful suggestions and remarks.  ... 
doi:10.1007/s10994-006-6888-8 fatcat:hixg2l3ulnhzhejhtbzmro6ohe

CASPER: A Case-Based Poker-Bot [chapter]

Ian Watson, Jonathan Rubin
2008 Lecture Notes in Computer Science  
CASPER improves upon previous case-based reasoning approaches to poker and is able to play evenly against the University of Alberta's Pokibots and Simbots, from which it acquired its case-bases and updates  ...  CASPER uses knowledge of previous poker scenarios to inform its betting decisions.  ...  A simulation-based betting strategy is analogous to selective search in perfect information games.  ... 
doi:10.1007/978-3-540-89378-3_60 fatcat:wugqe54i5vhi5c26yqaazyrzxu

Mechanistic analysis of the search behaviour of Caenorhabditis elegans

L. C. M. Salvador, F. Bartumeus, S. A. Levin, W. S. Ryu
2014 Journal of the Royal Society Interface  
that controls for area-restricted search behaviour, and (ii) a time-independent, 'intrinsic' strategy that reduces spatial oversampling and improves random encounter success.  ...  This behavioural transition shows that different reorientation behaviours are governed by two processes: (i) an environmentally informed 'extrinsic' strategy that is influenced by recent experience and  ...  Yet, animals have a way to hedge their bets on their world model by generating an efficient background search template.  ... 
doi:10.1098/rsif.2013.1092 pmid:24430127 pmcid:PMC3899880 fatcat:pp2jt3yx4vgmtje7w7f7whhidm

Is Institutional Evolution in Online Communities Driven by Selection or Stochasticity? [article]

Qiankun Zhong, Seth Frey, Martin Hilbert
2022 arXiv   pre-print
Disentangling the selective and stochastic components of social system change enables us to identify the key features to organizational development in the long run.  ...  But sometimes institutional change is due to stochastic drives including drift, path dependency, and blind imitation.  ...  Third, selective and stochastic forces can vary across time. Institutions that have been beneficial at the early times can end up reducing the growth speed (e.g., the lock-in effects).  ... 
arXiv:2204.12521v1 fatcat:qmqh5qniwfca7mylb4f6s2f6eu

Extremal Optimization Combined with LM Gradient Search for MLP Network Learning

Peng Chen, Yong-Zai Lu, Yu-Wang Chen
2010 International Journal of Computational Intelligence Systems  
Gradient search based neural network training algorithm may suffer from local optimum, poor generalization and slow convergence.  ...  Inheriting the advantages of the two approaches, the proposed "EO-LM" method can avoid local minima and improve MLP network learning performance in generalization capability and computation efficiency.  ...  , as an attempt to speed up the process of searching the local minimum.  ... 
doi:10.1080/18756891.2010.9727728 fatcat:u4qkiq27onaetf7hlr54oybzty

Hybridizing Adaptive Genetic Algorithm with Chaos Searching Technique for Numerical Optimization

Dongping Tian
2016 International Journal of Grid and Distributed Computing  
In addition, half of the total evolutionary generation is utilized as one of the decision conditions so as to speed up the convergent process.  ...  can find global optimal or the closer-to-optimal solutions and have faster search speed as well as higher convergence rate.  ...  [19] , an improved AGA has been proposed, which can get an effective trade-off between convergence speed and global optimum.  ... 
doi:10.14257/ijgdc.2016.9.2.12 fatcat:2aaxy32lunddbmnmod4gjuiz3m
« Previous Showing results 1 — 15 out of 1,635 results