Filters








2,044 Hits in 3.3 sec

Noisy Blackbox Optimization with Multi-Fidelity Queries: A Tree Search Approach [article]

Rajat Sen, Kirthevasan Kandasamy, Sanjay Shakkottai
2018 arXiv   pre-print
In this work, we incorporate the multi-fidelity setup in the powerful framework of noisy black-box optimization through tree-like hierarchical partitions.  ...  We study the problem of black-box optimization of a noisy function in the presence of low-cost approximations or fidelities, which is motivated by problems like hyper-parameter tuning.  ...  The main contributions of this paper are as follows: (i) We model multiple fidelities in the framework of black-box optimization of a noisy function, with hierarchical partitions in Section 3.  ... 
arXiv:1810.10482v1 fatcat:6ujzkggy5bccbdgtzr5pu7igni

Procrastinated Tree Search: Black-box Optimization with Delayed, Noisy, and Multi-Fidelity Feedback [article]

Junxiong Wang, Debabrota Basu, Immanuel Trummer
2022 arXiv   pre-print
We experimentally validate on multiple synthetic functions and hyperparameter tuning problems that PCTS outperforms the state-of-the-art black-box optimization methods for feedbacks with different noise  ...  In black-box optimization problems, we aim to maximize an unknown objective function, where the function is only accessible through feedbacks of an evaluation or simulation oracle.  ...  Bandits with delayed, aggregated anonymous feedback. Sen, R., Kandasamy, K., and Shakkottai, S. (2018). Multi-fidelity black-box optimization with hierarchical partitions.  ... 
arXiv:2110.07232v2 fatcat:aypluotofffe7mjntmwtxigmr4

Skew-resistant parallel processing of feature-extracting scientific user-defined functions

YongChul Kwon, Magdalena Balazinska, Bill Howe, Jerome Rolia
2010 Proceedings of the 1st ACM symposium on Cloud computing - SoCC '10  
At the heart of the SkewReduce system is an optimizer, parameterized by user-defined cost functions, that determines how best to partition the input data to minimize computational skew.  ...  These applications exhibit significant computational skew, where the runtime of different partitions depends on more than just input size and can therefore vary dramatically and unpredictably.  ...  We estimate the total runtime by running the black-box scheduling algorithm.  ... 
doi:10.1145/1807128.1807140 dblp:conf/cloud/KwonBHR10 fatcat:uycukwm6ebdlhffywcv7ypzdjy

Model Agnostic Multilevel Explanations [article]

Karthikeyan Natesan Ramamurthy, Bhanukiran Vinzamuri, Yunfeng Zhang, Amit Dhurandhar
2020 arXiv   pre-print
In recent years, post-hoc local instance-level and global dataset-level explainability of black-box models has received a lot of attention.  ...  We validate the effectiveness of the proposed technique based on two human studies -- one with experts and the other with non-expert users -- on real world datasets, and show that we produce high fidelity  ...  with different degrees of cohesion, while at the same time being fidel to the black-box model.  ... 
arXiv:2003.06005v1 fatcat:q5jhrbvnnje7fpwjnpurw3nnum

Aircraft Geometry and Meshing with Common Language Schema CPACS for Variable-Fidelity MDO Applications

Mengmeng Zhang, Aidan Jungo, Alessandro Gastaldi, Tomas Melin
2018 Aerospace (Basel)  
This paper discusses multi-fidelity aircraft geometry modeling and meshing with the common language schema CPACS.  ...  The CPACS-based multi-fidelity aerodynamic tools show a great consistency due to the one data-centric schema, and the automation of the progressive process can thus be implemented and realized with minimum  ...  Multi-fidelity aerodynamic modeling aims to cover the flight state parameter space of the entire flight envelope with an optimal distribution of computational resources.  ... 
doi:10.3390/aerospace5020047 fatcat:qjfi5wapmbc2fa5aftmw525g5i

A Hierarchical Multigrid Method for Oil Production Optimization

Steen Hørsholt, Hamid Nick, John Bagterp Jørgensen
2019 IFAC-PapersOnLine  
Furthermore, we present a hierarchical multigrid method for oil production optimization. The method utilizes a hierarchy of coarse level models based on the high-fidelity model.  ...  Furthermore, we present a hierarchical multigrid method for oil production optimization. The method utilizes a hierarchy of coarse level models based on the high-fidelity model.  ...  Thus we treat the reservoir simulator as a black-box function, S(ū; x 0 , θ).  ... 
doi:10.1016/j.ifacol.2019.06.110 fatcat:zauq57zeiraepicbif5aukkgn4

Model-based methods for continuous and discrete global optimization

Thomas Bartz-Beielstein, Martin Zaefferer
2017 Applied Soft Computing  
The use of surrogate models is a standard method to deal with complex, realworld optimization problems. The first surrogate models were applied to continuous optimization problems.  ...  The article concludes with a discussion of recent developments and challenges in both application domains. $ This is an extended version of the contribution [1]  ...  data to approximate black-box simulations.  ... 
doi:10.1016/j.asoc.2017.01.039 fatcat:ghmsan4pdjgalab5akbzgubco4

Metamodeling techniques for CPU-intensive simulation-based design optimization: a survey

Hanane Khatouri, Tariq Benamara, Piotr Breitkopf, Jean Demange
2022 Advanced Modeling and Simulation in Engineering Sciences  
Thus, the concept of multi-fidelity proposes to merge different levels of fidelity within a single model with controlled variance.  ...  This paper reviews the strategies that seek to improve surrogate-based optimization efficiency, including ROM, multi-fidelity metamodeling, and DOE enrichment strategies.  ...  Another review presents methods tackling high-dimensional black-box problems by [189] . This section presents some of the existing enrichment criteria applied to the SBO framework.  ... 
doi:10.1186/s40323-022-00214-y fatcat:4n7uucw445b5vcalrrxcrcwctq

Fast Network Simulation Through Approximation or

Charles W. Kazer, João Sedoc, Kelvin K.W. Ng, Vincent Liu, Lyle H. Ungar
2018 Proceedings of the 17th ACM Workshop on Hot Topics in Networks - HotNets '18  
In the limit, the rest of the network could be modeled as a single black box, but training that black box to approximate such a large collection of machines is not trivial.  ...  For example, Figure 2 and Figure 3 show how our system would replace the four switches of each approximated cluster with a single black box approximation.  ... 
doi:10.1145/3286062.3286083 dblp:conf/hotnets/KazerSNLU18 fatcat:qjav44lnyfh2jlwax7i46zel6a

Harnessing Low-Fidelity Data to Accelerate Bayesian Optimization via Posterior Regularization [article]

Bin Liu
2019 arXiv   pre-print
Bayesian optimization (BO) is a powerful paradigm for derivative-free global optimization of a black-box objective function (BOF) that is expensive to evaluate.  ...  However, the overhead of BO can still be prohibitive for problems with highly expensive function evaluations.  ...  Techniques such as hierarchical partitioning [2] , hierarchical modeling [3] and ensemble methods [4] , are used to incorporate multiple fidelities/cheap approximations of the BOF.  ... 
arXiv:1902.03740v5 fatcat:sro7envc4vcbhhds5aptjbw5ii

Hyperparameter Optimization [chapter]

Matthias Feurer, Frank Hutter
2019 Automated Machine Learning  
Since the high computational demand of many modern machine learning applications renders pure blackbox optimization extremely costly, we next focus on modern multi-fidelity methods that use (much) cheaper  ...  We first discuss blackbox function optimization methods based on model-free methods and Bayesian optimization.  ...  optimization and is used as a workbench for the yearly Black-Box Optimization Benchmarking (BBOB) challenge [11] .  ... 
doi:10.1007/978-3-030-05318-5_1 fatcat:kq2qut5lonhpdbnw35jswlgupa

Multi-Information Source Optimization [article]

Matthias Poloczek, Jialei Wang, Peter I. Frazier
2016 arXiv   pre-print
We consider Bayesian optimization of an expensive-to-evaluate black-box objective function, where we also have access to cheaper approximations of the objective.  ...  Its optimization decisions rely on a value of information analysis that extends the Knowledge Gradient factor to the setting of multiple information sources that vary in cost: each sampling decision maximizes  ...  The task of optimizing a single, expensive-to-evaluate black-box function has received a lot of attention.  ... 
arXiv:1603.00389v1 fatcat:vranz5c2hjdlbfb6462i4efpru

Bayesian Optimization for Iterative Learning [article]

Vu Nguyen and Sebastian Schulze and Michael A Osborne
2021 arXiv   pre-print
Our algorithm outperforms all existing baselines in identifying optimal hyperparameters in minimal time.  ...  In this paper, we present a Bayesian optimization (BO) approach which exploits the iterative structure of learning algorithms for efficient hyperparameter tuning.  ...  Multi-fidelity BO with continuous approximation (BOCA) [15] and hierarchical partition [33] extend this idea to continuous settings.  ... 
arXiv:1909.09593v5 fatcat:4kvp4aeilfdl7fiq6wfmt5toia

Frontmatter

Simon M. Lucas, Michael Mateas, Mike Preuss, Pieter Spronck, Julian Togelius, Michael Wagner
2013 Dagstuhl Publications  
(such as Hierarchical Task Networks, dynamic scripting, "black box" strategy selection, player modelling and learning approaches) to yield a new generation of search-based AI approaches for video games  ...  "Black Box" Strategy Selection An intuitive approach to applying adversarial search to video games is to start off with a collection of action scripts that are capable of playing entire games and then  ... 
doi:10.4230/dfu.vol6.12191.i dblp:conf/dagstuhl/X13b fatcat:i4isdb5w4fastcbnczbtsdulkm

Multi-fidelity Bayesian Optimization with Max-value Entropy Search and its parallelization [article]

Shion Takeno, Hitoshi Fukuoka, Yuhki Tsukada, Toshiyuki Koyama, Motoki Shiga, Ichiro Takeuchi, Masayuki Karasuyama
2020 arXiv   pre-print
Multi-fidelity Bayesian optimization (MFBO) accelerates BO by incorporating lower fidelity observations available with a lower sampling cost.  ...  We show that, in our multi-fidelity MES (MF-MES), most of additional computations, compared with usual MES, is reduced to analytical computations.  ...  Introduction Bayesian optimization (BO) is a popular machine-learning technique for the black-box optimization problem.  ... 
arXiv:1901.08275v2 fatcat:uvbwtjhp2ncovjv3n7qcnphgpq
« Previous Showing results 1 — 15 out of 2,044 results