892,838 Hits in 3.5 sec

In Search of Robust Measures of Generalization [article]

Gintare Karolina Dziugaite, Alexandre Drouin, Brady Neal, Nitarshan Rajkumar, Ethan Caballero, Linbo Wang, Ioannis Mitliagkas, Daniel M. Roy
2021 arXiv   pre-print
We argue that generalization measures should instead be evaluated within the framework of distributional robustness.  ...  Building on their study, we highlight where their proposed methods can obscure failures and successes of generalization measures in explaining generalization.  ...  In this work, we propose empirical methodology to aid in the search of a precise mathematical theory, allowing us to leverage large-scale empirical studies of generalization, like those in recent work  ... 
arXiv:2010.11924v2 fatcat:7wkjwvd6ordrtjoqtbophst7vm

Fast algorithm for robust template matching with m-estimators

Jiun-Hung Chen, Chu-Song Chen, Yong-Sheng Chen
2003 IEEE Transactions on Signal Processing  
We propose a particular image hierarchy called the -pyramid that can be exploited to generate a list of ascending lower bounds of the minimal matching errors when a nondecreasing robust error measure is  ...  This fast algorithm ensures finding the global minimum of the robust template matching problem in which a nondecreasing M-estimator serves as an error measure.  ...  In essence, a set of ascending lower bounds of the minimal matching error can be generated with our method as long as the robust error measure is nondecreasing.  ... 
doi:10.1109/tsp.2002.806551 fatcat:tliow2ekhvbubdyx6e646v2vli

Trade-Off between Performance and Robustness: An Evolutionary Multiobjective Approach [chapter]

Yaochu Jin, Bernhard Sendhoff
2003 Lecture Notes in Computer Science  
Unfortunately, only one solution can usually be obtained from one run of optimization using the existing methods for searching robust solutions.  ...  In real-world applications, it is often desired that a solution is not only of high performance, but also of strong robustness.  ...  The robustness measure is then applied in multiobjective optimization in Section 3 to generate a set of Pareto-optimal solutions.  ... 
doi:10.1007/3-540-36970-8_17 fatcat:hm57jfgexjbgrjjcon7vlt7nqi

Classification and Coverage-Based Falsification for Embedded Control Systems [chapter]

Arvind Adimoolam, Thao Dang, Alexandre Donzé, James Kapinski, Xiaoqing Jin
2017 Lecture Notes in Computer Science  
This falsification algorithm combines global and local search methods and uses a classification technique based on support vector machines to identify regions of the search space on which to focus effort  ...  While coverage metrics are often used to measure the exhaustiveness of this kind of testing approach for software systems, existing falsification approaches for CPS designs do not consider coverage for  ...  In the existing research on cyber-physical systems testing, the focus was generally on state-coverage measures, that is measures to characterize the portion of the state space covered by a test suite.  ... 
doi:10.1007/978-3-319-63387-9_24 fatcat:oudb7ggvtzd6zlg7cmojlxqdsq

Differential evolution for optimal grouping of condition monitoring signals of nuclear components [chapter]

P Baraldi, E Zio, F Di Maio, L Pappaglione, R Chevalier, R Seraoui
2011 Advances in Safety, Reliability and Risk Management  
The results of the grouping are evaluated with respect to the accuracy and robustness of the estimates of the monitored signals by the condition monitoring model developed on the optimal groups, and compared  ...  This behavior, which is found also in the pure DE search, is in this case more remarkable given that the robustness decreases of about 30%.  ...  Figure 6 . 6 Evolution of the best (x) and mean (.) robustness value at each generation.  ... 
doi:10.1201/b11433-59 fatcat:xseqz5yrfjdlnczq2hzcttvdqi

Evolving Robust Gene Regulatory Networks

Nasimul Noman, Taku Monjo, Pablo Moscato, Hitoshi Iba, Alberto de la Fuente
2015 PLoS ONE  
The automatic designing capability of GRN topology that can exhibit robust behavior can dramatically change the current practice in synthetic biology.  ...  The proposed framework was verified using two classic GRN behaviors: oscillation and bistability, although the framework is generalized for evolving other types of responses.  ...  and N = 3 using the quantitative measure of robustness in (5) .  ... 
doi:10.1371/journal.pone.0116258 pmid:25616055 pmcid:PMC4304830 fatcat:nzzzbfj5p5ctfmkdotjjf5w64a

Robustness, Data Analysis, and Statistical Modeling: The First 50 Years and Beyond

Erniel B. Barrios
2015 Communications for Statistical Applications and Methods  
We also present the current interest in robust statistics and conclude with suggestions on the possible future direction of this area for statistical science.  ...  From the pioneering work of Tukey, Huber, and Hampel that focused on robust location parameter estimation, we presented various generalizations of these estimation procedures that cover a wide variety  ...  The method that uses a hybrid of the forward search and the EM algorithm were able to generate robust estimates in the presence of outliers.  ... 
doi:10.5351/csam.2015.22.6.543 fatcat:ng7vp6xward2dilrwbhceq35ay

Robust Video Watermarking using Secret Sharing and Cuckoo Search Algorithm

Cuckoo search algorithm is an efficiently designed algorithm for optimization based on the behavior of blood parasitism of Cuckoo species.  ...  With said advantages, a novel method on video watermarking using Cuckoo search algorithm in DWT-SVD transform domain is proposed. SSIM, BER are used for fitness function in optimization function.  ...  In most of the scenarios, the invisibility feature of a watermark is measured using NC and PSNR and the robustness feature is measured using NC and BER.  ... 
doi:10.35940/ijitee.k2112.0981119 fatcat:nvpf2n3jajgw3kw5al52ejx3da

Robust Optimization for RCPSP Under Uncertainty

Hayet Mogaadi, Besma Fayech Chaar
2016 International Journal of Software Engineering & Applications  
The studied robustness consists in minimizing the worst-case performance, referred to as the min-max robustness objective, among a set of initial scenarios.  ...  The aim of the present article is to optimize the robustness objective for the Resource-Constrained Project scheduling Problem (RCPSP) dealing with activity duration uncertainty.  ...  Usually, the global performance is measured in terms of mean value, maximum deviation, etc.  ... 
doi:10.5121/ijsea.2016.7205 fatcat:inotzoo62vckblalxpbpyyrfi4

A Practical Approach for Robust and Flexible Vehicle Routing Using Metaheuristics and Monte Carlo Sampling

Kenneth Sörensen, Marc Sevaux
2009 Journal of Mathematical Modelling and Algorithms  
In this paper, we investigate how robust and flexible solutions of a number of stochastic variants of the capacitated vehicle routing problem can be obtained.  ...  It is also more flexible in the sense that adaptation of the approach to more complex problems can be easily done.  ...  In our MA|PM, the robust evaluation function is used to guide the search in the binary tournament selection process.  ... 
doi:10.1007/s10852-009-9113-5 fatcat:if3qohsd3jhtpabfr7domlajn4

Incorporating robustness into web ranking evaluation

Xin Li, Fan Li, Shihao Ji, Zhaohui Zheng, Yi Chang, Anlei Dong
2009 Proceeding of the 18th ACM conference on Information and knowledge management - CIKM '09  
In many Web search engines, a ranking function is selected for deployment mainly by comparing the relevance measurements over candidates.  ...  We argue that the ranking robustness can be effectively measured by taking into account the ranking score distribution across search results.  ...  How to evaluate the proposed metrics with respect to their measurements of robustness is a key issue in this study.  ... 
doi:10.1145/1645953.1646288 dblp:conf/cikm/LiLJZCD09 fatcat:yaonmbmhpjhbbdzrfyk2eq3nxi

Exploring the Hyperparameter Landscape of Adversarial Robustness [article]

Evelyn Duesterwald and Anupama Murthi and Ganesh Venkataraman and Mathieu Sinn and Deepak Vijaykeerthy
2019 arXiv   pre-print
In this paper, we explore some of the practical challenges of adversarial training.  ...  Based on these findings, we present a practical approach that leverages hyperparameter optimization techniques for tuning adversarial training to maximize robustness while keeping the loss in accuracy  ...  We conducted a robustness test that measures robustness as the accuracy of the model on adversarial examples generated using PGD.  ... 
arXiv:1905.03837v1 fatcat:ismc2lpm5zbodpvhkngupk5jz4

Generalized Time-Series Active Search With Kullback–Leibler Distance for Audio Fingerprinting

H. Lin, Z. Ou, X. Xiao
2006 IEEE Signal Processing Letters  
To address the resulting complexity, generalized time-series active search is proposed, which supports a wide variety of distance measures between two CCGMMs, including 1 , 2 , KL, etc.  ...  In this letter, a new audio fingerprinting approach is presented.  ...  ), KL distance measure, and generalized active search.  ... 
doi:10.1109/lsp.2006.874394 fatcat:ive2dvbbeje4bm5bgn7dvu3k6m

Robust multiobjective optimisation for fuzzy job shop problems

Juan José Palacios, Inés González-Rodríguez, Camino R. Vela, Jorge Puente
2017 Applied Soft Computing  
The experimental results also serve to analyse the goodness of the predictive robustness measure, in terms of its correlation with simulations of the empirical measure.  ...  To this end, we consider two measures of solution robustness: a predictive one, prior to the schedule execution, and an empirical one, measured at execution.  ...  In the following, we give a precise definition of a measure for each kind of robustness.  ... 
doi:10.1016/j.asoc.2016.07.004 fatcat:ie66bcuxmnfh3grtgoacc7f6kq

Multi-scenario Multi-objective Robust Optimization under Deep Uncertainty: A posteriori approach⋆,⋆⋆

Babooshka Shavazipour, Jan H. Kwakkel, Kaisa Miettinen
2021 Environmental Modelling & Software  
We also find that the lake problem is ill-suited for reflecting trade-offs in robust performance over the set of scenarios and Pareto optimality in any specific scenario, highlighting the need for novel  ...  issues, and enables the decision-makers to explore the trade-off between optimality/feasibility in any selected scenario and robustness across a broader range of scenarios.  ...  This research is related to the thematic research area Decision Analytics utilizing Causal Models and Multiobjective Optimization (DEMO, of the University of Jyvaskyla.  ... 
doi:10.1016/j.envsoft.2021.105134 fatcat:a2ekw6xftfdinefjwfqh5s7tni
« Previous Showing results 1 — 15 out of 892,838 results