Penalized maximum-likelihood estimation, the Baum-Welch algorithm, diagonal balancing of symmetric matrices and applications to training acoustic data

DOI10.1016/S0377-0427(00)00385-XzbMath0962.65008WikidataQ127724750 ScholiaQ127724750MaRDI QIDQ1576464

Charles A. Micchelli, Peder A. Olsen

Publication date: 31 May 2001

Published in: Journal of Computational and Applied Mathematics (Search for Journal in Brave)

zbMATH Keywords

algorithmsEM algorithmlarge-scale optimizationhigh-dimensional data\(K\)-means clusteringnonparametric multivariate density estimationpenalized maximum-likelihood estimationtrain speech datatraining acoustic vectors for speech recognition

Mathematics Subject Classification ID

Nonparametric regression and quantile regression (62G08) Applications of statistics in engineering and industry; control charts (62P30) Numerical mathematical programming methods (65K05) Large-scale problems in mathematical programming (90C06)

Related Items

A complete asymptotic analysis of an oscillation free nonlinear equation of Bessel type with a pole in the dependent variable, Approximation algorithms for homogeneous polynomial optimization with quadratic constraints

Cites Work

• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• Unnamed Item
• On pairs of multidimensional matrices
• Matrix scaling, entropy minimization, and conjugate duality. I: Existence conditions
• On the scaling of multidimensional matrices
• Scalings of matrices which have prespecified row sums and column sums via optimization
• Generalized scalings satisfying linear equations
• Matrices diagonally similar to a symmetric matrix
• On the estimation of a probability density function by the maximum penalized likelihood method
• Flows on graphs applied to diagonal similarity and diagonal equivalence for matrices
• Concerning nonnegative matrices and doubly stochastic matrices
• Scaling of matrices to achieve specified row and column sums
• On a theorem of P. Nowosad
• On the integral equation \(kf=1/f\) arising in a problem in communication
• The spectrum of a nonlinear operator associated with a matrix
• Estimation of a multivariate density
• The diagonal equivalence of a nonnegative matrix to a stochastic matrix
• The DAD theorem for symmetric non-negative matrices
• Remarks on Some Nonparametric Estimates of a Density Function
• Matrix Analysis
• Smoothing and Interpolation in a Convex Subset of a Hilbert Space
• Classifications of Nonnegative Matrices Using Diagonal Equivalence
• Cones, graphs and optimal scalings of matrices
• Characterizations of optimal scalings of matrices
• Maximum likelihood estimation for multivariate observations of Markov sources
• Smoothing and interpolation in a convex subset of a Hilbert space : II. The semi-norm case
• Scaling Matrices to Prescribed Row and Column Maxima
• Algorithms for Testing the Diagonal Similarity of Matrices and Related Problems
• A Relationship Between Arbitrary Positive Matrices and Doubly Stochastic Matrices
• An inequality with applications to statistical estimation for probabilistic functions of Markov processes and to a model for ecology
• A Maximization Technique Occurring in the Statistical Analysis of Probabilistic Functions of Markov Chains
• Nonparametric Roughness Penalties for Probability Densities
• On Estimation of a Probability Density Function and Mode