Filters








3,221 Hits in 3.2 sec

Classification with Ultrahigh-Dimensional Features [article]

Yanming Li, Hyokyoung Hong, Jian Kang, Kevin He, Ji Zhu, Yi Li
2016 arXiv   pre-print
Although much progress has been made in classification with high-dimensional features Fan_Fan:2008, JGuo:2010, CaiSun:2014, PRXu:2014, classification with ultrahigh-dimensional features, wherein the features  ...  This paper introduces a novel and computationally feasible multivariate screening and classification method for ultrahigh-dimensional data.  ...  ACKNOWLEDGEMENTS And this is an acknowledgements section with a heading that was produced by the \section* command. Thank you all for helping me writing this L A T E X sample file.  ... 
arXiv:1611.01541v1 fatcat:sathoccmxfcczdtjc53c2wmopy

Feature screening for ultrahigh dimensional binary data

Guoyu Guan, Na Shan, Jianhua Guo
2018 Statistics and its Interface  
With the rapid development of information technology, ultrahigh dimensional binary data have increased dramatically, for which feature screening has become a necessary step in real data analysis.  ...  Lastly, its outstanding performance is numerically confirmed on simulated data, and a real example of Chinese document classification is presented for illustration purpose.  ...  The aforementioned methods are all developed for dealing with ultrahigh dimensional continuous predictors.  ... 
doi:10.4310/sii.2018.v11.n1.a4 fatcat:fkzd2cwzkfet5aefcn6gtroeni

Multiclass Linear Discriminant Analysis with Ultrahigh‐Dimensional Features

Yanming Li, Hyokyoung G. Hong, Yi Li
2019 Biometrics  
Within the framework of Fisher's discriminant analysis, we propose a multiclass classification method which embeds variable screening for ultrahigh-dimensional predictors.  ...  Leveraging interfeature correlations, we show that the proposed linear classifier recovers informative features with probability tending to one and can asymptotically achieve a zero misclassification rate  ...  We propose an ultrahigh-dimensional multiclass classification method within the framework of Fisher's LDA.  ... 
doi:10.1111/biom.13065 pmid:31009070 pmcid:PMC6810714 fatcat:oy5kcjkbsvhvrmucziirqmyjda

Model-Free Feature Screening for Ultrahigh Dimensional Discriminant Analysis

Hengjian Cui, Runze Li, Wei Zhong
2015 Journal of the American Statistical Association  
This work is concerned with marginal sure independence feature screening for ultrahigh dimensional discriminant analysis. The response variable is categorical in discriminant analysis.  ...  Third, it allows the categorical response having a diverging number of classes in the order of O(n κ ) with some κ ≥ 0.  ...  Mai and Zou (2013) developed a sure feature screening procedures with ultrahigh dimensional predictors based on the Kolmogorov distance, but it is studied only for binary classification problems.  ... 
doi:10.1080/01621459.2014.920256 pmid:26392643 pmcid:PMC4574103 fatcat:qmgcivdb4rhnfg2qnoo27bqouu

A selective overview of feature screening for ultrahigh-dimensional data

JingYuan Liu, Wei Zhong, RunZe Li
2015 Science China Mathematics  
This has been called ultrahigh-dimensional data in the literature. This work aims to present a selective overview of feature screening procedures for ultrahigh-dimensional data.  ...  Feature selection and variable selection are fundamental for high-dimensional data analysis.  ...  [57] advocated model-free feature screening procedures for ultrahigh-dimensional data.  ... 
doi:10.1007/s11425-015-5062-9 pmid:26779257 pmcid:PMC4711389 fatcat:bu7msj4eyvgyffxzxm457ud7lq

Feature Screening for Ultrahigh Dimensional Categorical Data With Applications

Danyang Huang, Runze Li, Hansheng Wang
2014 Journal of business & economic statistics  
We propose a Pearson chi-square based feature screening procedure for categorical response with ultrahigh dimensional categorical covariates.  ...  Ultrahigh dimensional data with both categorical responses and categorical covariates are frequently encountered in the analysis of big data, for which feature screening has become an indispensable statistical  ...  This article aims to develop a feature screening procedure for multiclass classification with ultrahigh dimensional categorical predictors.  ... 
doi:10.1080/07350015.2013.863158 pmid:25328278 pmcid:PMC4197855 fatcat:txaxerznhjernhowo7wizct2du

Feature Screening for Ultrahigh Dimensional Categorical Data with Applications

Danyang Huang, Runze Li, Hansheng Wang
2013 Social Science Research Network  
We propose a Pearson chi-square based feature screening procedure for categorical response with ultrahigh dimensional categorical covariates.  ...  Ultrahigh dimensional data with both categorical responses and categorical covariates are frequently encountered in the analysis of big data, for which feature screening has become an indispensable statistical  ...  This article aims to develop a feature screening procedure for multiclass classification with ultrahigh dimensional categorical predictors.  ... 
doi:10.2139/ssrn.2378670 fatcat:647uenokvjaujh5nvucubkfghu

Exploring the microstructure manifold: image texture representations applied to ultrahigh carbon steel microstructures [article]

Brian L. DeCost and Toby Francis and Elizabeth A. Holm
2017 arXiv   pre-print
high-dimensional microstructure representations.  ...  We introduce a microstructure informatics dataset focusing on complex, hierarchical structures found in a single Ultrahigh carbon steel under a range of heat treatments.  ...  Data visualization We used t-SNE to visualize the distributions of high-dimensional microstructure features obtained with each feature extraction method. 2 t-SNE is an unsupervised technique; the only  ... 
arXiv:1702.01117v2 fatcat:4mhijwlb2zbefjj6vwj6wqpbbq

Ultrahigh-Dimensional Multiclass Linear Discriminant Analysis by Pairwise Sure Independence Screening

Rui Pan, Hansheng Wang, Runze Li
2016 Journal of the American Statistical Association  
This article is concerned with the problem of feature screening for multiclass linear discriminant analysis under ultrahigh-dimensional setting. We allow the number of classes to be relatively large.  ...  To solve the problem, we propose a novel pairwise sure independence screening method for linear discriminant analysis with an ultrahigh-dimensional predictor.  ...  As a result, document classification can be formulated as a classification problem with ultrahigh-dimensional feature and a large number of classes.  ... 
doi:10.1080/01621459.2014.998760 pmid:28127109 pmcid:PMC5256914 fatcat:iijrwcxkevdapcqo4httptzi5e

Making Decision Trees Feasible in Ultrahigh Feature and Label Dimensions

Weiwei Liu, Ivor W. Tsang
2017 Journal of machine learning research  
Extensive empirical studies verify that 1) SBT is easy to understand and interpret in ultrahigh dimensions and is more resilient to noisy features. 2) Compared with state-of-the-art algorithms, our proposed  ...  sparse coding tree framework is more efficient, yet accurate in ultrahigh label and feature dimensions.  ...  This paper first studies two emerging challenges for decision trees -ultrahigh feature dimensionality and ultrahigh label dimensionality.  ... 
dblp:journals/jmlr/LiuT17 fatcat:f3n3l7yf3fatlhxc4itqdhlowy

Ultrahigh Dimensional Feature Screening via RKHS Embeddings

Krishnakumar Balasubramanian, Bharath K. Sriperumbudur, Guy Lebanon
2013 International Conference on Artificial Intelligence and Statistics  
Feature screening is a key step in handling ultrahigh dimensional data sets that are ubiquitous in modern statistical problems.  ...  ., Lasso/sparse additive model) have been extensively developed and analyzed for feature selection in high dimensional regime.  ...  in ultrahigh dimensional settings.  ... 
dblp:conf/aistats/BalasubramanianSL13 fatcat:gfepgupmybc73e5jsvmphxbjki

Weighted Mean Squared Deviation Feature Screening for Binary Features

Gaizhen Wang, Guoyu Guan
2020 Entropy  
In this study, we propose a novel model-free feature screening method for ultrahigh dimensional binary features of binary classification, called weighted mean squared deviation (WMSD).  ...  Compared to Chi-square statistic and mutual information, WMSD provides more opportunities to the binary features with probabilities near 0.5.  ...  In this study, we focus on feature screening of binary classification with ultrahigh dimensional binary features.  ... 
doi:10.3390/e22030335 pmid:33286109 pmcid:PMC7516793 fatcat:5l37z4xfxjaq5lszqqpkrw7sqm

Learning Representations of Ultrahigh-dimensional Data for Random Distance-based Outlier Detection

Guansong Pang, Longbing Cao, Ling Chen, Huan Liu
2018 Proceedings of the 24th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining - KDD '18  
Learning expressive low-dimensional representations of ultrahigh-dimensional data, e.g., data with thousands/millions of features, has been a major way to enable learning methods to address the curse of  ...  Extensive empirical results on eight real-world ultrahigh dimensional data sets show that REPEN (i) enables a random distance-based detector to obtain significantly better AUC performance and two orders  ...  However, both approaches are mainly focused on data sets with tens/hundreds of features due to their prohibitive subspace search in the ultrahigh-dimensional space.  ... 
doi:10.1145/3219819.3220042 dblp:conf/kdd/PangCCL18 fatcat:mm5yxsxkxjabheri4oczr2bgka

Deep Learning Approaches to Image Texture Analysis in Material Processing

Xiu Liu, Chris Aldrich
2022 Metals  
In the simulated case studies, material microstructures were simulated with Voronoi graphic representations and in the real-world case study, the appearance of ultrahigh carbon steel is cast as a textural  ...  The ability of random forest models, as well as the convolutional neural networks themselves, to discriminate between different textures with the image features as input was used as the basis for comparison  ...  original high-dimensional space with those in the low-dimensional space.  ... 
doi:10.3390/met12020355 fatcat:tmdhtmbfejgtpl73dtywvllbwa

HIGH DIMENSIONALITY REDUCTION ON GRAPHICAL DATA

Smita J.Khelukar .
2015 International Journal of Research in Engineering and Technology  
This is especially serious for high dimensional data with little examples.  ...  To meet this test, a novel effective structure to perform highlight determination for graph embedding, in which a classification of graph implanting routines is given a role as a slightest squares relapse  ...  During learning process, basic gathering structures of related components connected with every support feature indicated as Affiliated features can likewise be found with no extra cost.  ... 
doi:10.15623/ijret.2015.0411029 fatcat:of6zgaorlvfmfednwebtxni23q
« Previous Showing results 1 — 15 out of 3,221 results