Filters








88,291 Hits in 4.2 sec

Adaptation Algorithm and Theory Based on Generalized Discrepancy [article]

Corinna Cortes and Mehryar Mohri and Andres Muñoz Medina
2015 arXiv   pre-print
We present a new algorithm for domain adaptation improving upon a discrepancy minimization algorithm previously shown to outperform a number of algorithms for this task.  ...  We show that our algorithm benefits from a solid theoretical foundation and more favorable learning bounds than discrepancy minimization.  ...  in much of learning theory and algorithmic design.  ... 
arXiv:1405.1503v3 fatcat:pwldig74rnamhdwv3yqdmayzne

Adaptation Algorithm and Theory Based on Generalized Discrepancy

Corinna Cortes, Mehryar Mohri, Andrés Muñoz Medina
2015 Proceedings of the 21th ACM SIGKDD International Conference on Knowledge Discovery and Data Mining - KDD '15  
Instead, it uses a reweighting that depends on the hypothesis considered and is based on the minimization of a new measure of generalized discrepancy.  ...  We present a new algorithm for domain adaptation improving upon the discrepancy minimization algorithm (DM), which was previously shown to outperform a number of popular algorithms designed for this task  ...  We present an alternative theoretically well founded algorithm for domain adaptation that is based on minimizing a finer quantity, the generalized discrepancy, and that seeks to improve upon DM.  ... 
doi:10.1145/2783258.2783368 dblp:conf/kdd/CortesMM15 fatcat:uiuvopyfxzcpdodtbtdp4q2vdm

Bridging Theory and Algorithm for Domain Adaptation [article]

Yuchen Zhang, Tianle Liu, Mingsheng Long, Michael I. Jordan
2019 arXiv   pre-print
Existing domain adaptation theories naturally imply minimax optimization algorithms, which connect well with the domain adaptation methods based on adversarial learning.  ...  We extend previous theories (Mansour et al., 2009c; Ben-David et al., 2010) to multiclass classification in domain adaptation, where classifiers based on the scoring functions and margin loss are standard  ...  Conclusion We present a theoretical and algorithmic analysis of domain adaptation, which bridges the gap between theory and algorithm.  ... 
arXiv:1904.05801v2 fatcat:5tn7padlb5glpcosn5kt7ezooe

Discrepancy-based Evolutionary Diversity Optimization [article]

Aneta Neumann, Wanru Gao, Carola Doerr, Frank Neumann, Markus Wagner
2018 arXiv   pre-print
Our experimental investigations comparing three diversity optimization approaches show that a discrepancy-based diversity optimization approach using a tie-breaking rule based on weighted differences to  ...  Diversity optimization in terms of features on the underlying problem allows to obtain a better understanding of possible solutions to the problem at hand and can be used for algorithm selection when dealing  ...  All of the experiments are executed on a machine with 48-core AMD 2.80 GHz CPU and 128 GByte RAM. Each algorithm is run for 20 000 generations and the final discrepancy is reported.  ... 
arXiv:1802.05448v1 fatcat:nj3byono6fe4jjroeg27lfjb6e

Metric Information Matrix for Maximum Mean Discrepancy for Domain Adaptation

Wenjuan Ren, Shie Zhou, Zhanpeng Yang, Quan Shi, Xian Sun, Luyi Yang
2021 IEEE Access  
Based on Gromov-Hausdorff's theory, we proposed two kinds of feature mappings in the model of joint distribution adaptation by embedding the original feature subspace to a common subspace.  ...  Our experiments show that constructed mappings have the abilities to alleviate the feature discrepancy and mitigate the distribution shift between source domain and target domains.  ...  Generally speaking, for any N ≤ |X |, X and M N (X ) cannot be placed in one-to-one correspondence. If N = |X |, then MMD algorithms based on labels and pseudo-labels doesn't work on this subspace.  ... 
doi:10.1109/access.2021.3123281 fatcat:ddpaitylfnhw3cqmbzowenv4q4

An Interpolation Procedure for List Decoding Reed–Solomon Codes Based on Generalized Key Equations

Alexander Zeh, Christian Gentner, Daniel Augot
2011 IEEE Transactions on Information Theory  
This article provides a link between syndrome-based decoding approaches based on Key Equations and the interpolation-based list decoding algorithms of Guruswami and Sudan for Reed-Solomon codes.  ...  List decoding algorithms, capable of decoding beyond half the minimum distance, are based on interpolation and factorization of multivariate polynomials.  ...  Sidorenko and M. Bossert for fruitful discussions. They thank the anonymous referees for their valuable comments that improved the presentation of this paper.  ... 
doi:10.1109/tit.2011.2162160 fatcat:2ipefyqnsvavvjffk6wwsc75sq

Parallel Quasi-Monte Carlo Integration by Partitioning Low Discrepancy Sequences [chapter]

Alexander Keller, Leonhard Grünschloß
2012 Monte Carlo and Quasi-Monte Carlo Methods 2010  
Independent of the number of processors, the resulting algorithms are deterministic, and generalize and improve upon previous approaches.  ...  The approach in fact partitions a single low-discrepancy sequence into multiple low-discrepancy sequences.  ...  In the following a strictly deterministic scheme is introduced: Based on a generalized and simplified argument on how to partition quadrature rules for parallel processing, efficient algorithms for generating  ... 
doi:10.1007/978-3-642-27440-4_27 fatcat:xwjvnhnb7ne5vaubvaky4dffda

Towards Accurate and Robust Domain Adaptation under Noisy Environments [article]

Zhongyi Han, Xian-Jin Gui, Chaoran Cui, Yilong Yin
2020 arXiv   pre-print
Previous domain adaptation works have achieved great success in theory and practice.  ...  To reduce the impact of feature noise, we propose a proxy distribution based margin discrepancy.  ...  Adaptation (ADDA) [Tzeng et al., 2017], Margin Disparity Discrepancy based algorithm (MDD) [Zhang et al., 2019], and Transferable Curriculum Learning (TCL) Figure 4 : 4 The t-SNE visualization of class  ... 
arXiv:2004.12529v2 fatcat:uu2z5oslyjbsrml4waefuoiddq

Editorial

Camil Demetrescu, Magnus M. Halldorsson
2013 Algorithmica  
Discrepancy of a set system is a problem that has for long admitted only existential bounds based on the probabilistic method.  ...  The European Symposium on Algorithms is one of the premier algorithms conferences worldwide, featuring high-quality papers on efficient algorithms and data structures in computer science, discrete applied  ...  Together these six beautiful articles show the breadth and the depth of algorithms research in general, and of the European Symposium on Algorithms in particular.  ... 
doi:10.1007/s00453-013-9824-x fatcat:6r5uldd7brb7vjbkkon26iltvm

A Theory of Multiple-Source Adaptation with Limited Target Labeled Data [article]

Yishay Mansour and Mehryar Mohri and Jae Ro and Ananda Theertha Suresh and Ke Wu
2020 arXiv   pre-print
We show that a new family of algorithms based on model selection ideas benefits from very favorable guarantees in this scenario and discuss some theoretical obstacles affecting some alternative techniques  ...  We present a theoretical and algorithmic study of the multiple-source domain adaptation problem in the common scenario where the learner has access only to a limited amount of labeled target data, but  ...  However, such an algorithm would incur higher generalization errors and due to space constraints, we discuss it in Appendix D.1. We now discuss two algorithms based on discrepancy.  ... 
arXiv:2007.09762v2 fatcat:m2jlolwjsndnpgznxrt4dmuiwi

Discrepancy-based evolutionary diversity optimization

Aneta Neumann, Wanru Gao, Carola Doerr, Frank Neumann, Markus Wagner
2018 Proceedings of the Genetic and Evolutionary Computation Conference on - GECCO '18  
Our experimental investigations comparing three diversity optimization approaches show that a discrepancy-based diversity optimization approach using a tiebreaking rule based on weighted differences to  ...  Diversity optimization in terms of features on the underlying problem allows to obtain a better understanding of possible solutions to the problem at hand and can be used for algorithm selection when dealing  ...  Acknowledgment This work has been supported through Australian Research Council (ARC) grants DP140103400 and DP160102401.  ... 
doi:10.1145/3205455.3205532 dblp:conf/gecco/NeumannGDN018 fatcat:atsgpnsjuzd5rekwgvo7j7usai

A Combined Prediction Model for Subgrade Settlement Based on Improved Set Pair Analysis

Yafeng Li, Wuming Leng, Rusong Nie, Huihao Mei, Siwei Zhou, Junli Dong
2019 Mathematical Problems in Engineering  
Then the risk of the set pair relationship identification was expressed based on the Bayesian decision theory, and the optimal criterion of set pair relationship was obtained by the adaptive search algorithm  ...  of the single prediction model and make the combined prediction based on the certainty degree, and the criterion of set pair relationship was optimized.  ...  Acknowledgments The authors' research is supported by the National Natural Science Foundation of China (nos. 51878666 and 51678572) and the Innovation Project of Central South University (Grant no. 2018zzts192  ... 
doi:10.1155/2019/1064246 fatcat:i77jbtzwkfhxzjkzqj2qnbv33m

Smoothed Analysis with Adaptive Adversaries [article]

Nika Haghtalab, Tim Roughgarden, Abhishek Shetty
2021 arXiv   pre-print
Crucially, our results hold for adaptive adversaries that can choose an input distribution based on the decisions of the algorithm and the realizations of the inputs in the previous time steps.  ...  -Online discrepancy minimization: We consider the online Komlós problem, where the input is generated from an adaptive sequence of σ-smooth and isotropic distributions on the ℓ_2 unit ball.  ...  distributions up front, an adaptive adversary can base its current choice of an input distribution on the decisions of the algorithm and the realizations of the inputs in previous time steps.  ... 
arXiv:2102.08446v2 fatcat:eq3326qer5aohmxnqmirqc3wcq

A Weighted Partial Domain Adaptation for Acoustic Scene Classification and Its Application in Fiber Optic Security System

Ningyu He, Jie Zhu
2020 IEEE Access  
General domain adaptation assume that the source and the target domain have the same label space.  ...  Experiments are carried out on TUT and ESC-50 datasets which show that our method achieves state-of-the-art results.  ...  NETWORK FRAMEWORK The weighted partial domain adaptation we proposed is based on the theory of generative adversarial neural networks (GAN) [19] .  ... 
doi:10.1109/access.2020.3044153 fatcat:smrtlxa6gbbbfi5w7i3w6kf2ai

Page 2087 of Mathematical Reviews Vol. , Issue 2002C [page]

2002 Mathematical Reviews  
This monograph on the discrepancy method describes how the tools of discrepancy theory have been successfully used to gain insight into algorithm design and complexity theory.  ...  The versatility of the evolutionary procedure is one of its main strengths in serving as a basis for generating intelligent behavior, adapting to new challenges, and learning from experience.  ... 
« Previous Showing results 1 — 15 out of 88,291 results