Large-sample properties of unsupervised estimation of the linear discriminant using projection pursuit
From MaRDI portal
Publication:2074341
Abstract: We study the estimation of the linear discriminant with projection pursuit, a method that is blind in the sense that it does not use the class labels in the estimation. Our viewpoint is asymptotic and, as our main contribution, we derive central limit theorems for estimators based on three different projection indices, skewness, kurtosis and their convex combination. The results show that in each case the limiting covariance matrix is proportional to that of linear discriminant analysis (LDA), an unblind estimator of the discriminant. An extensive comparative study between the asymptotic variances reveals that projection pursuit is able to achieve efficiency equal to LDA when the groups are arbitrarily well-separated and their sizes are reasonably balanced. We conclude with a real data example and a simulation study investigating the validity of the obtained asymptotic formulas for finite samples.
Recommendations
Cites work
- A generalized projection pursuit procedure and its significance level
- An alternative to EM for Gaussian mixture models: batch and stochastic Riemannian optimization
- Asymptotic Theory for Principal Component Analysis
- ASYMPTOTIC THEORY FOR PRINCIPAL COMPONENT ANALYSIS: NON-NORMAL CASE1
- CHIME: clustering of high-dimensional Gaussian mixtures with EM algorithm and its optimality
- Cluster Identification Using Projections
- Efficiently learning mixtures of two Gaussians
- Eigenvectors of a kurtosis matrix as interesting directions to reveal cluster structure
- Exploratory Projection Pursuit
- Fourth moments and independent component analysis
- scientific article; zbMATH DE number 1964693 (Why is no real title available?)
- Identification of Outliers in Multivariate Data
- Invariant co-ordinate selection (with discussion)
- Limit distributions for measures of multivariate skewness and kurtosis based on projections
- Maximum likelihood estimation of Gaussian mixture models using stochastic search
- Model-Based Clustering, Discriminant Analysis, and Density Estimation
- Notion of information and independent component analysis.
- On the concentration of the missing mass
- Principal component analysis.
- Projection pursuit
- Projection pursuit in high dimensions
- REPPlab: an R package for detecting clusters and outliers using exploratory projection pursuit
- Significance levels in exploratory projection pursuit
- Skewness and the linear discriminant function
- Skewness-based projection pursuit: a computational approach
- Some theory for Fisher's linear discriminant function, `naive Bayes', and some alternatives when there are many more variables than observations
- Spectral norm of a symmetric tensor and its computation
- Statistical Analysis of Financial Data in S-Plus
- Statistical guarantees for the EM algorithm: from population to sample-based analysis
- The Deflation-Based FastICA Estimator: Statistical Analysis Revisited
- The elements of statistical learning. Data mining, inference, and prediction
- The stochastic EM algorithm: Estimation and asymptotic results
- The tube method for the moment index in projection pursuit
- The use of a common location measure in the invariant coordinate selection and projection pursuit
- Tight bounds for learning a mixture of two Gaussians (extended abstract)
- Two statistics for testing for multivariate normality
- Vector-valued skewness for model-based clustering
- What is Projection Pursuit?
Cited in
(3)
Describes a project that uses
Uses Software
This page was built for publication: Large-sample properties of unsupervised estimation of the linear discriminant using projection pursuit
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2074341)
