代码搜索结果

function p = gaussian_prob(x, m, C, use_log) % GAUSSIAN_PROB Evaluate a multivariate Gaussian density. % p = gaussian_prob(X, m, C) % p(i) = N(X(:,i), m, C) where C = covariance matrix and each COL

function Y = sample_cond_multinomial(X, M) % SAMPLE_MULTINOMIAL Sample Y(i) ~ M(X(i), :) % function Y = sample_multinomial(X, M) % % X(i) = i'th sample % M(i,j) = P(Y=j | X=i) = noisy channel mod

function M = sample_gaussian(mu, Sigma, N) % SAMPLE_GAUSSIAN Draw N random row vectors from a Gaussian distribution % samples = sample_gaussian(mean, cov, N) if nargin==2 N = 1; end % If Y

% Verify that my code gives the same results as the 1D example at % http://www.media.mit.edu/physics/publications/books/nmm/files/cwm.m seed = 0; rand('state', seed); randn('state', seed); x =

function [S, mu, sigma2] = standardize(M, mu, sigma2) % function S = standardize(M, mu, sigma2) % Make each column of M be zero mean, std 1. % Thus each row is scaled separately. % % If mu, sigma

function [S, mu, sigma2] = standardize(M, mu, sigma2) % function S = standardize(M, mu, sigma2) % Make each column of M be zero mean, std 1. % Thus each row is scaled separately. % % If mu, sigma

function [mu, Sigma, weights] = mixgauss_init(M, data, cov_type, method) % MIXGAUSS_INIT Initial parameter estimates for a mixture of Gaussians % function [mu, Sigma, weights] = mixgauss_init(M, dat

function test_eval_pdf_cond_mixgauss() %Q = 10; M = 100; d = 20; T = 500; Q = 2; M = 3; d = 4; T = 5; mu = rand(d,Q,M); data = randn(d,T); %mixmat = mk_stochastic(rand(Q,M)); mixmat = mk_sto

function r = betarnd(a,b,m,n); %BETARND Random matrices from beta distribution. % R = BETARND(A,B) returns a matrix of random numbers chosen % from the beta distribution with parameters A an

function p = matrix_T_pdf(A, M, V, K, n) % MATRIX_T_PDF Evaluate the density of a matrix under a Matrix-T distribution % p = matrix_T_pdf(A, M, V, K, n) % See "Bayesian Linear Regression", T. Min