Filters








457 Hits in 3.5 sec

Unsupervised Domain Adaptation with Copula Models [article]

Cuong D. Tran and Ognjen Rudovic and Vladimir Pavlovic
2017 arXiv   pre-print
We study the task of unsupervised domain adaptation, where no labeled data from the target domain is provided during training time.  ...  By transforming the data to a copula domain, we show on a number of benchmark datasets (including human emotion estimation), and using different regression models for prediction, that we can achieve a  ...  To avoid the requirement for labeled target data, unsupervised domain adaptation with feature-transformation is employed.  ... 
arXiv:1710.00018v1 fatcat:olb7ypflarg57phqdijg46amhu

Semi-Supervised Domain Adaptation with Non-Parametric Copulas [article]

David Lopez-Paz, José Miguel Hernández-Lobato, Bernhard Schölkopf
2013 arXiv   pre-print
A new framework based on the theory of copulas is proposed to address semi- supervised domain adaptation problems.  ...  Therefore, changes in each of these factors can be detected and corrected to adapt a density model accross different learning domains.  ...  Comparison with other Domain Adaptation Methods NPRV is analyzed in a series of experiments for domain adaptation on the non-linear regression setting with real-world data.  ... 
arXiv:1301.0142v1 fatcat:jyni2boikjhntosqwsxcux5ltq

Table of Contents

2021 IEEE transactions on fuzzy systems  
Pedrycz 3293 Multisource Heterogeneous Unsupervised Domain Adaptation via Fuzzy Relation Neural Networks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  ...  Pedrycz 3532 SHORT PAPERS Extensions of Discrete Copulas to Sparse Copulas . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .  ... 
doi:10.1109/tfuzz.2021.3119193 fatcat:htsh37qgo5gxpgb77kbvhv6lbe

Benchmarking Unsupervised Outlier Detection with Realistic Synthetic Data [article]

Georg Steinbuss, Klemens Böhm
2020 arXiv   pre-print
Benchmarking unsupervised outlier detection is difficult. Outliers are rare, and existing benchmark data contains outliers with various and unknown characteristics.  ...  This might be due to the imprecise notion of outliers or to the difficulty to arrive at a good coverage of different domains with synthetic data.  ...  [34] make use of a vine copula for outlier detection. Clearly, comparing other detection methods to one based on a vine copula can be biased when the data is generated from a vine copula.  ... 
arXiv:2004.06947v1 fatcat:fhqubsxnzjattkuywhzju25o4u

Supervised Classification of Very High Resolution Optical Images Using Wavelet-Based Textural Features

Olivier Regniers, Lionel Bombrun, Virginie Lafon, Christian Germain
2016 IEEE Transactions on Geoscience and Remote Sensing  
Its high adaptability and the low number of parameters to be set are other advantages of the proposed approach.  ...  A strategy is proposed to apply these models in a supervised classification framework.  ...  [12] suggested an unsupervised segmentation approach in which each vine plot is isolated by selecting its frequency response in the Fourier domain using adapted Gabor filters.  ... 
doi:10.1109/tgrs.2016.2526078 fatcat:cgv2lpz3e5a5lif7v3zrpl3u7e

Unsupervised segmentation of randomly switching data hidden with non-Gaussian correlated noise

Pierre Lanchantin, Jérôme Lapuyade-Lahorgue, Wojciech Pieczynski
2011 Signal Processing  
The interest of the proposed models and related processing is validated by different experiments some of which are related to semi-supervised and unsupervised image segmentation.  ...  The aim of this paper is to propose a simultaneous solution to both of these problems using triplet Markov chains (TMC) and copulas.  ...  Fig. 5 contains the results of unsupervised segmentation of the data Y =y, once with the true Student copula and once with the wrong Gaussian copula.  ... 
doi:10.1016/j.sigpro.2010.05.033 fatcat:ath3kzjakjbklhdkhryhhpvvqe

Incorporating Regular Vines in Estimation of Distribution Algorithms [chapter]

Rogelio Salinas-Gutiérrez, Arturo Hernández-Aguirre, Enrique R. Villa-Diharce
2013 Studies in Computational Intelligence  
Moreover, this chapter also shows how the use of mutual information in the learning of graphical models implies a natural way of employing copula functions.  ...  Copula theory has been used also for modeling multivariate distributions in unsupervised learning problems such as image segmentation [11, 19] and retrieval tasks [39, 49, 58] .  ...  Moreover, vines can be easily adapted to higher dimensions.  ... 
doi:10.1007/978-3-642-32726-1_2 fatcat:2msakluycfcilnosf6auqq4l4a

Data-SUITE: Data-centric identification of in-distribution incongruous examples [article]

Nabeel Seedat, Jonathan Crabbé, Mihaela van der Schaar
2022 arXiv   pre-print
by a model trained with the training instances?  ...  DATA-SUITE leverages copula modeling, representation learning, and conformal prediction to build feature-wise confidence interval estimators based on a set of training instances.  ...  Mind the class weight bias: Weighted maximum mean discrepancy for unsupervised domain adaptation.  ... 
arXiv:2202.08836v2 fatcat:vh7tzoy6jzcprc643osxo6qtce

Wavelet Domain Astronomical Multiband Image Fusion And Restoration Using Markov Quadtree And Copulas

Farid Flitti, Christophe Collet, Eric Slezak
2005 Zenodo  
Model parameter estimation The quadtree defined in section 3 with a likelihood model described in subsection 3.1 needs a parameter estimation procedure to be unsupervised.  ...  On the other hand, wavelet framework is very well adapted for denoising task [2] . Thus wavelet domain seems a quite appropriate for noisy image fusion.  ... 
doi:10.5281/zenodo.39312 fatcat:ewpfy6hx7zg25njqmxs2yeeoeq

Wheezing sounds detection using multivariate generalized gaussian distributions

S. Le Cam, A. Belghith, Ch. Collet, F. Salzenstein
2009 2009 IEEE International Conference on Acoustics, Speech and Signal Processing  
We suggest a modeling for wheezing and normal sounds in the wavelet packet domain using generalized gaussian distributions.  ...  We cope with the multidimensional aspect of the generalized gaussian distribution by using the theory of copulas. Experimental results are given in detail in this paper.  ...  This project is done in collaboration with the "Hôpitaux Universitaire de Strasbourg".  ... 
doi:10.1109/icassp.2009.4959640 dblp:conf/icassp/CamBCS09 fatcat:v2wpesd25fhmxpjaas4stk4uky

Mixture Kernel Density Estimation and Remedied Correlation Matrix on the EEG-Based Copula Model for the Assessment of Visual Discomfort

Yawen Zheng, Xiaojie Zhao, Li Yao
2020 Cognitive Computation  
With the favourable quality of the proposed EEG-based model, it is used to extract time-domain EEG features to assess visual discomfort further.  ...  Although the copula model can explore the dependence among variables, the EEG-based copula models still have the following deficiencies: (1) the methods ignoring the fine-grained information hidden in  ...  The AR model is also adopted to extract time-domain EEG features to compare with the feature 2C_M_R.  ... 
doi:10.1007/s12559-020-09780-y fatcat:nicsxqjcibaappivox2l5fryqm

Clustering Using Student t Mixture Copulas

Till Massing
2021 SN Computer Science  
(Parametric characterization of multimodal distributions with non-Gaussian modes, pp 286–292, 2011) introduced Gaussian mixture copula models (GMCM) for clustering problems which do not assume normality  ...  In this paper, we propose Student t mixture copula models (SMCM) as an extension of GMCMs. GMCMs require weak assumptions, yielding a flexible fit and a powerful cluster tool.  ...  Acknowledgements Financial support of the German Research Foundation (Deutsche Forschungsgemeinschaft, DFG) via the Collaborative Research Center "Statistical modelling of nonlinear dynamic processes"  ... 
doi:10.1007/s42979-021-00503-0 fatcat:nxnlimzkzva2ndpdfsubce42rm

Copulas as High-Dimensional Generative Models: Vine Copula Autoencoders [article]

Natasa Tagasovska, Damien Ackerer, Thibault Vatter
2019 arXiv   pre-print
Second, the multivariate distribution of the encoded data is estimated with vine copulas.  ...  Third, a generative model is obtained by combining the estimated distribution with the decoder part of the AE.  ...  Semi-supervised domain adaptation with copulas. NeurIPS, 2013. [48] Alireza Makhzani, Jonathon Shlens, Navdeep Jaitly, and Ian Goodfellow. Adversarial autoen- coders. In ICLR, 2016.  ... 
arXiv:1906.05423v2 fatcat:ggjandqzz5gszmahdrzqsdijhi

An Unsupervised Hierarchical Rule Based Model for Aspect Term Extraction Augmented with Pruning Strategies

Manju Venugopalan, Deepa Gupta
2020 Procedia Computer Science  
The proposed model has reported an appreciable recall of 81.9 and 68.7 on Restaurant and Laptop domains respectively on SemEval 2014 dataset and has also been compared with the state of art models.  ...  The proposed model has reported an appreciable recall of 81.9 and 68.7 on Restaurant and Laptop domains respectively on SemEval 2014 dataset and has also been compared with the state of art models.  ...  The model performance has been compared with three unsupervised/hybrid baselines reported on the same dataset.  ... 
doi:10.1016/j.procs.2020.04.303 fatcat:zg3mqxjtlfbbdbgjxixtryxbcu

Survey of State-of-the-Art Mixed Data Clustering Algorithms

Amir Ahmad, Shehroz Khan
2019 IEEE Access  
We analyze the strengths and weaknesses of these methods with pointers for future research directions.  ...  Mixed data comprises both numeric and categorical features, and mixed datasets occur frequently in many domains, such as health, finance, and marketing.  ...  Rajan and Bhattacharya [59] Gaussian mixture copula. Tekumalla1 et al. [60] Vine copulas and Dirichlet process mixture of vines. Marbac [61] A mixture model of Gaussian copulas. Foss et al.  ... 
doi:10.1109/access.2019.2903568 fatcat:f34d2e7tvfbqjpz26srrsworuu
« Previous Showing results 1 — 15 out of 457 results