Filters








24,949 Hits in 5.1 sec

Subspace Analysis Using Random Mixture Models

Xiaogang Wang, Xiaoou Tang
2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition (CVPR'05)  
In this paper, we develop a random mixture model to improve Bayes and LDA subspace analysis.  ...  Subspace Analysis Based on a Single Gaussian Model We first briefly review several conventional subspace analysis methods, including Bayes, LDA, and null space LDA.  ...  Discussion Our subspace analysis using random mixture models can be understood as a multiple classifier integration framework.  ... 
doi:10.1109/cvpr.2005.336 dblp:conf/cvpr/WangT05 fatcat:oxol2dfwk5ghdav4jcm3apdije

Random Subspace Mixture Models for Interpretable Anomaly Detection [article]

Cetin Savkli, Catherine Schwartz
2021 arXiv   pre-print
Gaussian mixture models (GMMs) are used to create the probability densities for each subspace with techniques included to mitigate singularities allowing for the ability to handle both numerical and categorial  ...  In selecting random subspaces, equal representation of each attribute is used to ensure correct statistical limits.  ...  To accomplish these goals, we present a new model, Random Subspace Mixture Model (RSMM), which leverages some of the useful features of existing methods.  ... 
arXiv:2108.06283v1 fatcat:3i3kkndmu5c67ibaz2jsxzu5qm

Multivariate calibration of non-replicated measurements for the factored noise model

Nirav Bhatt, Shankar Narasimhan
2009 Chemometrics and Intelligent Laboratory Systems  
As a first step, an iterative procedure is developed to estimate the error variance factors along with the spectral subspace, which is subsequently used in developing the regression model.  ...  Using simulated and experimental data, it is shown that the quality of the MVC model developed using the proposed method is better than that obtained using PCR, and is as good as the model obtained using  ...  In the first step, principal component analysis (PCA) is used to estimate a lower dimension subspace from the spectroscopic data.  ... 
doi:10.1016/j.chemolab.2009.06.005 fatcat:2ngrmm5darh4lfxpziwdlrsi2u

An algorithm for separation of mixed sparse and Gaussian sources

Ameya Akkalkotkar, Kevin Scott Brown, Boris Podobnik
2017 PLoS ONE  
For mixtures of purely Gaussian sources, principal component analysis (PCA) can provide a basis for the Gaussian subspace.  ...  These multiple decompositions allow us to assess component stability as the size of the data matrix changes, which can be used to determinine the dimension of the nongaussian subspace in a mixture.  ...  We therefore continued to use simulated mixtures, but the nongaussian subspace was constructed from random sections of the public domain audiobooks described in Methods.  ... 
doi:10.1371/journal.pone.0175775 pmid:28414814 pmcid:PMC5393591 fatcat:wvnt54wq5nfvznh7oo4func3iy

Multivariate calibration of non-replicated measurements for heteroscedastic errors

Nirav P. Bhatt, Amit Mitna, Shankar Narasimhan
2007 Chemometrics and Intelligent Laboratory Systems  
to the accuracy of the model obtained using MLPCR.  ...  The core of the proposed approach is an Iterative Principal Component Analysis method which simultaneously estimates the lower dimensional spectral subspace and all the error variances.  ...  Introduction Multivariate calibration (MVC) methods are routinely used for analysis of chemical mixtures.  ... 
doi:10.1016/j.chemolab.2006.04.006 fatcat:l7sr37xvfzagbeutffdxjakhpm

Learning Low-Dimensional Signal Models

Lawrence Carin, Richard Baraniuk, Volkan Cevher, David Dunson, Michael Jordan, Guillermo Sapiro, Michael Wakin
2011 IEEE Signal Processing Magazine  
The mixture-of-factor-analyzer (MFA) model discussed below is a statistical form of the union-of-subspace data model, and the MFA may also be used to approximate a manifold.  ...  In this analysis the incomplete data (image) are used as model inputs, to infer all model parameters, and importantly A and {c n • b n } n=1,N .  ... 
doi:10.1109/msp.2010.939733 pmid:24363544 pmcid:PMC3869461 fatcat:xwhiardgrrde3c5267f7nowb7u

A linear discriminant analysis framework based on random subspace for face recognition

Xiaoxun Zhang, Yunde Jia
2007 Pattern Recognition  
However, it remains a problem how to construct an optimal random subspace for discriminant analysis and perform the most efficient discriminant analysis on the constructed random subspace.  ...  Random subspace can effectively solve this problem by random sampling on face features.  ...  Wang and Tang [4] used random subspace to en force Fisherface and they also developed a random mixture model to improve LDA subspace analysis in [15] .  ... 
doi:10.1016/j.patcog.2006.12.002 fatcat:rpfbezfzxrfbfa4mnirwaelg3a

Page 5387 of Mathematical Reviews Vol. , Issue 88j [page]

1988 Mathematical Reviews  
The interesting feature of this mixture is that it is a data- adaptive mixture, so that the estimator will shrink mainly to the subspace (or subspaces) most consistent with the data.  ...  Random-effect models together with a variety of computer-intensive iterative techniques have been suggested for the analysis of such data.  ... 

Subspace K-means clustering

Marieke E. Timmerman, Eva Ceulemans, Kim De Roover, Karla Van Leeuwen
2013 Behavior Research Methods  
Using data from a study on parental behavior, we show that subspace K-means analysis provides a rich insight into the cluster characteristics, in terms of both the relative positions of the clusters (via  ...  mixtures of factor analyzers (MFA), and MCLUST.  ...  Subspace K-means simultaneously models the centroids and the within-cluster residuals in subspaces, using a component analysis approach.  ... 
doi:10.3758/s13428-013-0329-y pmid:23526258 fatcat:n7qgv2nmcbgofj3pe2vze2bn6y

SOURCE SEPARATION IN THE PRESENCE OF SIDE INFORMATION: NECESSARY AND SUFFICIENT CONDITIONS FOR RELIABLE DE-MIXING

Zahra Sabetsarvestani, Francesco Renna, Franz Kiraly, Miguel R. D. Rodrigues
2018 2018 IEEE Global Conference on Signal and Information Processing (GlobalSIP)  
By positing that the individual components of the mixed signals as well as the corresponding side information signals follow a joint Gaussian mixture model, we characterise necessary and sufficient conditions  ...  number of measurements from the mixture, then we can reliably separate the sources, otherwise we cannot.  ...  In this model m represent the number of random measurements extracted from the linear mixture, where we assume m < n x .  ... 
doi:10.1109/globalsip.2018.8646499 dblp:conf/globalsip/Sabetsarvestani18 fatcat:sbmbkjisnvaqxfgwck7dt3peye

Probabilistic model-based discriminant analysis and clustering methods in chemometrics

Charles Bouveyron
2013 Journal of Chemometrics  
Model-based techniques for discriminant analysis and clustering are popular tools which are renowned for their probabilistic foundations and their flexibility.  ...  The use of these model-based methods is also illustrated on real-world classification problems in Chemometrics using R packages. c (θ; y, z) = n i=1 K k=1 z ik log (π k f (y i ; θ k )) ,  ...  The MFA model is an extension of the factor analysis model to a mixture of K factor analyzers. Let {y 1 , . . . , y n } be independent observed realizations of a random vector Y ∈ R p .  ... 
doi:10.1002/cem.2563 fatcat:bcowqw3z2zbtdd3r75gacgeuku

Probabilistic model-based discriminant analysis and clustering methods in chemometrics

Charles Bouveyron
2013 Journal of Chemometrics  
Model-based techniques for discriminant analysis and clustering are popular tools which are renowned for their probabilistic foundations and their flexibility.  ...  The use of these model-based methods is also illustrated on real-world classification problems in Chemometrics using R packages. c (θ; y, z) = n i=1 K k=1 z ik log (π k f (y i ; θ k )) ,  ...  The MFA model is an extension of the factor analysis model to a mixture of K factor analyzers. Let {y 1 , . . . , y n } be independent observed realizations of a random vector Y ∈ R p .  ... 
doi:10.1002/cem.2560 fatcat:yyxavwgv6vhrjgse3yje2nqti4

An investigation into subspace rapid speaker adaptation for verification

S. Lucey, Tsuhan Chen
2003 2003 International Conference on Multimedia and Expo. ICME '03. Proceedings (Cat. No.03TH8698)  
should be used to construct the adaptation parametric subspace.  ...  Results are presented on the acoustic portion of the XM2VTS database for the task of Gaussian mixture model (GMM) based text-independent speaker verification.  ...  In this paper we have investigated the possible use of linear discriminant analysis [4] (LDA), which uses a criterion of class separation to define its subspace, to restrict a client's parametric representation  ... 
doi:10.1109/icme.2003.1220856 dblp:conf/icmcs/LuceyC03 fatcat:jm6xfvdf7bbfbokgugnkwxc2g4

Towards a general independent subspace analysis

Fabian J. Theis
2006 Neural Information Processing Systems  
By introducing the concept of irreducible independent subspaces or components, we present a generalization to a parameter-free mixture model.  ...  However, a serious drawback of k-ISA (and hence of ICA) lies in the fact that the requirement fixed group-size k does not allow us to apply this analysis to an arbitrary random vector. Indeed,  ...  Conclusion Previous approaches for independent subspace analysis were restricted either to fixed group sizes or semi-parametric models.  ... 
dblp:conf/nips/Theis06 fatcat:i4hcuiw5tfblbkxgy2v7z2jsmi

On Two Distinct Sources of Nonidentifiability in Latent Position Random Graph Models [article]

Joshua Agterberg, Minh Tang, Carey E. Priebe
2020 arXiv   pre-print
In this paper we define and examine these two nonidentifiabilities, dubbed subspace nonidentifiability and model-based nonidentifiability, in the context of random graph inference.  ...  Then, we characterize the limit for model-based nonidentifiability both with and without subspace nonidentifiability.  ...  analysis.  ... 
arXiv:2003.14250v1 fatcat:sskeukiffjcvlmpmjyxppu4x7q
« Previous Showing results 1 — 15 out of 24,949 results