38,872 Hits in 4.9 sec

Personalizing Performance Regression Models to Black-Box Optimization Problems [article]

Tome Eftimov, Anja Jankovic, Gorjan Popovski, Carola Doerr, Peter Korošec
2021 pre-print
With this work, we bring to the attention of our community the possibility to personalize regression models to specific types of optimization problems.  ...  Going one step further, we also investigate the impact of selecting not a single regression model per problem, but personalized ensembles.  ...  We also thank Pascal Kerschke, University of Dresden, for providing us with the flacco tool [23] used to compute the feature values for the 24 BBOB functions.  ... 
doi:10.1145/3449639.3459407 arXiv:2104.10999v1 fatcat:tz4wsn5yfzgy3b5ocjrp2niwum

Contemporary Symbolic Regression Methods and their Relative Performance [article]

William La Cava, Patryk Orzechowski, Bogdan Burlacu, Fabrício Olivetti de França, Marco Virgolin, Ying Jin, Michael Kommenda, Jason H. Moore
2021 arXiv   pre-print
We assess 14 symbolic regression methods and 7 machine learning methods on a set of 252 diverse regression problems.  ...  For the real-world datasets, we benchmark the ability of each method to learn models with low error and low complexity relative to state-of-the-art machine learning methods.  ...  The authors would also like to thank James McDermott for his generous contributions to the repository, and Randal Olson and Weixuan Fu for their initial push to integrate regression benchmarking into PMLB  ... 
arXiv:2107.14351v1 fatcat:ggzwkdhkunhcxps7p6spbliuha

Accuracy vs. robustness: Bi-criteria optimized ensemble of metamodels

Can Cui, Teresa Wu, Mengqi Hu, Jeffery D. Weir, Xianghua Chu
2014 Proceedings of the Winter Simulation Conference 2014  
Extensive research has investigated the performance of different metamodeling techniques in terms of accuracy and/or robustness and concluded no model outperforms others across diverse problem structures  ...  Regression and Radial Basis Function), where uncertainties are modeled for evaluating robustness.  ...  The Pareto optimality plots of most of black-box problems show that the two objectives are conflictive.  ... 
doi:10.1109/wsc.2014.7019926 dblp:conf/wsc/CuiWHWC14 fatcat:n6t2mua7lradpkf6a546744eba

Radial Basis Surrogate Model Integrated To Evolutionary Algorithm For Solving Computation Intensive Black-Box Problems

Abdulbaset Saad, Adel Younis, Zuomin Dong
2017 Zenodo  
This work proposes a series of modification to the Differential Evolution (DE) algorithm for solving computation Intensive Black-Box Problems.  ...  A meta-modeling assisted DE is proposed to solve computationally expensive optimization problems.  ...  The main focus of this modification was to increase the performance of DE algorithm on expensive black-box problems.  ... 
doi:10.5281/zenodo.1128937 fatcat:mjiqytdhgnhvvoz6y267uxtmlu

Parallelized bayesian optimization for problems with expensive evaluation functions

Margarita Rebolledo, Frederik Rehbach, A. E. Eiben, Thomas Bartz-Beielstein
2020 Proceedings of the 2020 Genetic and Evolutionary Computation Conference Companion  
Many black-box optimization problems rely on simulations to evaluate the quality of candidate solutions. These evaluations can be computationally expensive and very time-consuming.  ...  Each method is evaluated on all the 24 objective functions of the Black-Box-Optimization-Benchmarking test suite in their 20-dimensional versions.  ...  CONCLUSIONS To answer our research questions the performance of different parallel-BO implementations was compared against CMA-ES and random search on a simulated expensive black box optimization problem  ... 
doi:10.1145/3377929.3390017 dblp:conf/gecco/CoyREB20 fatcat:r2ikwtifn5etpc2yy3z23cuehm

Adversarial Attacks and Defenses in Physiological Computing: A Systematic Review [article]

Dongrui Wu, Weili Fang, Yi Zhang, Liuqing Yang, Xiaodong Xu, Hanbin Luo, Xiang Yu
2021 arXiv   pre-print
the training and/or test examples to hijack the machine learning algorithm output, leading to possibly user confusion, frustration, injury, or even death.  ...  However, the vulnerability of physiological computing systems has not been paid enough attention to, and there does not exist a comprehensive review on adversarial attacks to it.  ...  model in black-box attacks.  ... 
arXiv:2102.02729v3 fatcat:p2mqt3owajahbn5k6dukoiarau

Black-Box Optimization of Object Detector Scales [article]

Mohandass Muthuraja and Octavio Arriaga and Paul Plöger and Frank Kirchner and Matias Valdenegro-Toro
2020 arXiv   pre-print
We also perform a regression analysis to find the significant hyper-parameters to tune.  ...  In this work, we propose the use of Black-box optimization methods to tune the prior/default box scales in Faster R-CNN and SSD, using Bayesian Optimization, SMAC, and CMA-ES.  ...  We also show that the scales learned with black-box optimization transfer from PASCAL VOC 2007 to VOC 2012, with an mAP improvement as well, and we perform a regression analysis to find out which are the  ... 
arXiv:2010.15823v1 fatcat:ab3yxoigxrcu3isrg42hhmr33i

Taxonomic Multi-class Prediction and Person Layout Using Efficient Structured Ranking [chapter]

Arpit Mittal, Matthew B. Blaschko, Andrew Zisserman, Philip H. S. Torr
2012 Lecture Notes in Computer Science  
In this case we would like the mis-classification score to be proportional to the number of parts misclassified. It transpires both of these are examples of structured output ranking problems.  ...  Another example in vision is for the ubiquitous pictorial structure or parts based model.  ...  box of the person.  ... 
doi:10.1007/978-3-642-33709-3_18 fatcat:fhjofi63jzeyjco2vizbrwknhu

Text-based Person Search in Full Images via Semantic-Driven Proposal Generation [article]

Shizhou Zhang, Duo Long, Yitao Gao, Liying Gao, Qian Zhang, Kai Niu, Yanning Zhang
2021 arXiv   pre-print
To close the gap, we study the problem of text-based person search in full images by proposing a new end-to-end learning framework which jointly optimize the pedestrian detection, identification and visual-semantic  ...  Besides, a cross-scale visual-semantic embedding mechanism is utilized to improve the performance.  ...  feature embedding tasks together and jointly optimize them to achieve better performance.  ... 
arXiv:2109.12965v1 fatcat:xws4cmklxbh4pjwtbubxt5cqgu

Robust Adversarial Perturbation on Deep Proposal-based Models [article]

Yuezun Li, Daniel Tian, Ming-Ching Chang, Xiao Bian, Siwei Lyu
2019 arXiv   pre-print
Our method focuses on attacking the common component in these algorithms, namely Region Proposal Network (RPN), to universally degrade their performance in a black-box fashion.  ...  To do so, we design a loss function that combines a label loss and a novel shape loss, and optimize it with respect to image using a gradient based iterative algorithm.  ...  As such, our R-AP method is suitable in nature for black-box attack to these models, i.e., without the need to know their implementation details.  ... 
arXiv:1809.05962v2 fatcat:njrzt7sixjculij267wmehkhie

Stop Explaining Black Box Machine Learning Models for High Stakes Decisions and Use Interpretable Models Instead [article]

Cynthia Rudin
2019 arXiv   pre-print
People have hoped that creating methods for explaining these black box models will alleviate some of these problems, but trying to explain black box models, rather than creating models that are interpretable  ...  Black box machine learning models are currently being used for high stakes decision-making throughout society, causing problems throughout healthcare, criminal justice, and in other domains.  ...  , and several others who helped my thought processes in various ways, and particularly Berk Ustun, Ron Parr, Rob Holte, and my father, Stephen Rudin, who went to considerable efforts to provide thoughtful  ... 
arXiv:1811.10154v3 fatcat:2xqiy3n4irczza67iiuvxyrt7a

CorrAttack: Black-box Adversarial Attack with Structured Search [article]

Zhichao Huang, Yaowei Huang, Tong Zhang
2020 arXiv   pre-print
The time-varying contextual bandits problem can then be solved by a Bayesian optimization procedure, which can take advantage of the features of the structured action space.  ...  We show that searching over the structured space can be approximated by a time-varying contextual bandits problem, where the attacker takes feature of the associated arm to make modifications of the input  ...  Therefore, the attacker may query the black-box model and perform zeroth order optimization to find an adversarial example without the knowledge of the target model.  ... 
arXiv:2010.01250v1 fatcat:notqcsgqvzdn5eqyb4v54xalwy

An Analysis of Adversarial Attacks and Defenses on Autonomous Driving Models [article]

Yao Deng, Xi Zheng, Tianyi Zhang, Chen Chen, Guannan Lou, Miryung Kim
2020 arXiv   pre-print
However, it is uncertain to what extent regression models such as driving models are vulnerable to adversarial attacks, the effectiveness of existing defense techniques, and the defense implications for  ...  a good coverage of various attacks, (2) a blackbox attack is much less effective compared with a white-box attack, implying that it is important to keep model details (e.g., model architecture, hyperparameters  ...  Based on the inputs and outputs from the target model, to perform a black-box attack, attackers can build a substitute model and achieve white-box attacks on their own model.  ... 
arXiv:2002.02175v1 fatcat:v5xwnfd5rfglfcdrosld3zbvke

PRETZEL: Opening the Black Box of Machine Learning Prediction Serving Systems [article]

Yunseong Lee, Alberto Scolari, Byung-Gon Chun, Marco Domenico Santambrogio, Markus Weimer, Matteo Interlandi
2018 arXiv   pre-print
Current prediction serving systems consider models as black boxes, whereby prediction-time-specific optimizations are ignored in favor of ease of deployment.  ...  In this paper, we present PRETZEL, a prediction serving system introducing a novel white box architecture enabling both end-to-end and multi-model optimizations.  ...  Scenarios: The goals of our experimental evaluation are to evaluate how the white box approach performs compared to black box.  ... 
arXiv:1810.06115v1 fatcat:wnckhrapeja57nc6gciunsyq2e

Explainable Landscape-Aware Optimization Performance Prediction [article]

Risto Trajanov and Stefan Dimeski and Martin Popovski and Peter Korošec and Tome Eftimov
2021 arXiv   pre-print
In this study, we are investigating explainable landscape-aware regression models where the contribution of each landscape feature to the prediction of the optimization algorithm performance is estimated  ...  Efficient solving of an unseen optimization problem is related to appropriate selection of an optimization algorithm and its hyper-parameters.  ...  ACKNOWLEDGMENT We thank Diederick Vermetten, Leiden University, for providing us the modular CMA-ES performance data.  ... 
arXiv:2110.11633v1 fatcat:zk7dricidjfibap4syjqosad64
« Previous Showing results 1 — 15 out of 38,872 results