Search PubMed⌕ Search

SEARCH · Search PubMed

Results for “EM algorithm”

Search indexed PubMed citations on genomics, clinical trials, systematic reviews and public health. Explore titles, authors and supplied subject terms, then open the PubMed record.

Quote a phrase for an exact phrase match. Source license links do not imply unrestricted reuse.

At least 127 records · Page 7Linked to original sources
PubMed1998

Compression of strings with approximate repeats.

We describe a model for strings of characters that is loosely based on the Lempel Ziv model with the addition that a repeated substring can be an approximate match to the original substring; this is close to the situation of DNA, for example. Typically there are many explanations for a given string under the model, some optimal and many suboptimal. Rather than commit to one optimal explanation, we sum the probabilities over all explanations under the model because this gives the probability of the data under the model. The model has a small number of parameters and these can be estimated from the given string by an expectation-maximization (EM) algorithm. Each iteration of the EM algorithm takes O(n2) time and a few iterations are typically sufficient. O(n2) complexity is impractical for strings of more than a few tens of thousands of characters and a faster approximation algorithm is also given. The model is further extended to include approximate reverse complementary repeats when analyzing DNA strings. Tests include the recovery of parameter estimates from known sources and applications to real DNA strings.

Algorithms↗
PubMed1999

Independent factor analysis.

We introduce the independent factor analysis (IFA) method for recovering independent hidden sources from their observed mixtures. IFA generalizes and unifies ordinary factor analysis (FA), principal component analysis (PCA), and independent component analysis (ICA), and can handle not only square noiseless mixing but also the general case where the number of mixtures differs from the number of sources and the data are noisy. IFA is a two-step procedure. In the first step, the source densities, mixing matrix, and noise covariance are estimated from the observed data by maximum likelihood. For this purpose we present an expectation-maximization (EM) algorithm, which performs unsupervised learning of an associated probabilistic model of the mixing situation. Each source in our model is described by a mixture of gaussians; thus, all the probabilistic calculations can be performed analytically. In the second step, the sources are reconstructed from the observed data by an optimal nonlinear estimator. A variational approximation of this algorithm is derived for cases with a large number of sources, where the exact algorithm becomes intractable. Our IFA algorithm reduces to the one for ordinary FA when the sources become gaussian, and to an EM algorithm for PCA in the zero-noise limit. We derive an additional EM algorithm specifically for noiseless IFA. This algorithm is shown to be superior to ICA since it can learn arbitrary source densities from the data. Beyond blind separation, IFA can be used for modeling multidimensional data by a highly constrained mixture of gaussians and as a tool for nonlinear signal encoding.

Algorithms↗
PubMed2005

Expectation maximization algorithm based haplotype relative risk (EM-HRR): test of linkage disequilibrium using incomplete case-parents trios.

The Haplotype Relative Risk (HRR) was first proposed [Falk et al., Ann Hum Genet 1987] to test for Linkage Disequilibrium (LD) between a marker and a putative disease locus using case-parent trios. Spurious association does not appear in such family-based studies under population admixture. In this paper, we extend the HRR to accommodate incomplete trios via the Expectation-Maximization (EM) algorithm [Dempster et al., J R Stat Soc Ser B, 1977]. In addition to triads and dyads (parent-offspring pair), the EM-HRR easily incorporates individuals with no parental genotype information available, which is excluded from the one parent Transmission/Disequilibrium Test (1-TDT) [Sun et al., Am J Epidemiol 1999]. Due to the data structure of EM-HRR, transmitted alleles are always available regardless of the number of missing parental genotypes. As a result of having a larger sample size, computer simulations reveal that the EM-HRR is more powerful in detecting LD than the 1-TDT in a population under Hardy-Weinberg Equilibirum (HWE). If admixture is not extreme, the EM-HRR remains more powerful. When a large degree of admixture exists, the EM-HRR performs better the 1-TDT when the association is strong, though not as well when the association is weak. We illustrate the proposed method with an application to the Framingham Heart Study.

Algorithms↗
PubMed1992

Computer-assisted analysis of mixtures (C.A.MAM): statistical algorithms.

This paper presents various algorithmic approaches for computing the maximum likelihood estimator of the mixing distribution of a one-parameter family of densities and provides a unifying computer-oriented concept for the statistical analysis of unobserved heterogeneity (i.e., observations stemming from different subpopulations) in a univariate sample. The case with unknown number of population subgroups as well as the case with known number of population subgroups, with emphasis on the first, is considered in the computer package C.A.MAN (Computer Assisted Mixture Analysis). It includes an algorithmic menu with choices of the EM algorithm, the vertex exchange algorithm, a combination of both, as well as the vertex direction method. To ensure reliable convergence, a step-length menu is provided for the three latter methods, each achieving monotonicity for the direction of choice. C.A.MAN has the option to work with restricted support size-that is, the case when the number of components is known a priori. In the latter case, the EM algorithm is used. Applications of mixture modelling in medical problems are discussed.

Algorithms↗
PubMed1995

Maximum-likelihood estimation of molecular haplotype frequencies in a diploid population.

Molecular techniques allow the survey of a large number of linked polymorphic loci in random samples from diploid populations. However, the gametic phase of haplotypes is usually unknown when diploid individuals are heterozygous at more than one locus. To overcome this difficulty, we implement an expectation-maximization (EM) algorithm leading to maximum-likelihood estimates of molecular haplotype frequencies under the assumption of Hardy-Weinberg proportions. The performance of the algorithm is evaluated for simulated data representing both DNA sequences and highly polymorphic loci with different levels of recombination. As expected, the EM algorithm is found to perform best for large samples, regardless of recombination rates among loci. To ensure finding the global maximum likelihood estimate, the EM algorithm should be started from several initial conditions. The present approach appears to be useful for the analysis of nuclear DNA sequences or highly variable loci. Although the algorithm, in principle, can accommodate an arbitrary number of loci, there are practical limitations because the computing time grows exponentially with the number of polymorphic loci. Although the algorithm, in principle, can accommodate an arbitrary number of loci, there are practical limitations because the computing time grows exponentially with the number of polymorphic loci.

Algorithms↗
PubMed2004

Notes on the maximum likelihood estimation of haplotype frequencies.

The maximum likelihood estimation (MLE) is one of the most popular ways to estimate haplotype frequencies of a population with genotype data whose linkage phases are unknown. The MLE is commonly implemented in the use of the Expectation-Maximization (EM) algorithm. It is known that the EM algorithm carries the risk that an estimator may converge erroneously to one of the local maxima or saddle points of the likelihood surface, resulting in serious errors in the MLE of haplotype frequencies. In this note, by theoretical treatments we present the necessary and sufficient conditions that the local maxima or saddle points on the likelihood surface appear. As a rule of thumb, that the difference between the coupling and repulsive haplotype frequencies in phase known individuals is 3/2 times larger than the frequency of phase ambiguous individuals is the sufficient condition that the likelihood surface is unimodal. Moreover, we present the analytic solution to the biallelic two-locus problem, and construct a general algorithm to obtain the global maximum.

Algorithms↗
PubMed1999

Regression modeling of ordinal data with nonzero baselines.

This paper develops regression models for ordinal data with nonzero control response probabilities. The models are especially useful in dose-response studies where the spontaneous or natural response rate is nonnegligible and the dosage is logarithmic. These models generalize Abbott's formula, which has been commonly used to model binary data with nonzero background observations. We describe a biologically plausible latent structure and develop an EM algorithm for fitting the models. The EM algorithm can be implemented using standard software for ordinal regression. A toxicology data set where the proposed model fits the data but a more conventional model fails is used to illustrate the methodology.

Algorithms↗
PubMed2004

A fast algorithm for functional mapping of complex traits.

By integrating the underlying developmental mechanisms for the phenotypic formation of traits into a mapping framework, functional mapping has emerged as an important statistical approach for mapping complex traits. In this note, we explore the feasibility of using the simplex algorithm as an alternative to solve the mixture-based likelihood for functional mapping of complex traits. The results from the simplex algorithm are consistent with those from the traditional EM algorithm, but the simplex algorithm has considerably reduced computational times. Moreover, because of its nonderivative nature and easy implementation with current software, the simplex algorithm enjoys an advantage over the EM algorithm in the dynamic modeling and analysis of complex traits.

Algorithms↗
PubMedDate not supplied

Non-ignorable missing covariates in generalized linear models.

We propose a likelihood method for estimating parameters in generalized linear models with missing covariates and a non-ignorable missing data mechanism. In this paper, we focus on one missing covariate. We use a logistic model for the probability that the covariate is missing, and allow this probability to depend on the incomplete covariate. We allow the covariates, including the incomplete covariate, to be either categorical or continuous. We propose an EM algorithm in this case. For a missing categorical covariate, we derive a closed form expression for the E- and M-steps of the EM algorithm for obtaining the maximum likelihood estimates (MLEs). For a missing continuous covariate, we use a Monte Carlo version of the EM algorithm to obtain the MLEs via the Gibbs sampler. The methodology is illustrated using an example from a breast cancer clinical trial in which time to disease progression is the outcome, and the incomplete covariate is a quality of life physical well-being score taken after the start of therapy. This score may be missing because the patients are sicker, so this covariate could be non-ignorably missing.

Algorithms↗
PubMed1993

Evaluating interventions with differential attrition: the importance of nonresponse mechanisms and use of follow-up data.

Evaluations of psychological interventions are often criticized because of differential attrition, which is cited as a severe threat to validity. The present study shows that differential attrition is not a problem unless the mechanism causing the attrition is inaccessible (unavailable for analysis). With a simulation study, we show that conclusions about program effects (a) are unbiased when there is no differential attrition, even with usual complete cases analysis; (b) may be severely biased when based on usual complete cases analyses and there is differential attrition; (c) are unbiased when based on the expectation-maximization (EM) algorithm, even when there is differential attrition, as long as the attrition mechanism is accessible; and (d) are biased, even with the EM algorithm, when the attrition mechanism is inaccessible. Following Little and Rubin (1987), we advocate the collection of new data from a random sample of subjects with initially missing data. On the basis of these data, we propose a simple correction to the EM algorithm estimates. In our study, the correction produced unbiased estimates of program effects parameters, even with an inaccessible attrition mechanism and substantial differential attrition.

Bias↗
PubMed1999

Likelihood methods for incomplete longitudinal binary responses with incomplete categorical covariates.

We consider longitudinal studies in which the outcome observed over time is binary and the covariates of interest are categorical. With no missing responses or covariates, one specifies a multinomial model for the responses given the covariates and uses maximum likelihood to estimate the parameters. Unfortunately, incomplete data in the responses and covariates are a common occurrence in longitudinal studies. Here we assume the missing data are missing at random (Rubin, 1976, Biometrika 63, 581-592). Since all of the missing data (responses and covariates) are categorical, a useful technique for obtaining maximum likelihood parameter estimates is the EM algorithm by the method of weights proposed in Ibrahim (1990, Journal of the American Statistical Association 85, 765-769). In using the EM algorithm with missing responses and covariates, one specifies the joint distribution of the responses and covariates. Here we consider the parameters of the covariate distribution as a nuisance. In data sets where the percentage of missing data is high, the estimates of the nuisance parameters can lead to highly unstable estimates of the parameters of interest. We propose a conditional model for the covariate distribution that has several modeling advantages for the EM algorithm and provides a reduction in the number of nuisance parameters, thus providing more stable estimates in finite samples.

Affect↗
PubMed1992

Three-dimensional SPECT reconstruction of combined cone beam and parallel beam data.

Single photon emission computed tomography (SPECT) using cone beam (CB) collimation exhibits increased sensitivity compared with acquisition geometries using parallel (P) hole collimation. However, CB collimation has a smaller field-of-view which may result in truncated projections and image artifacts. A primary objective of this work is to investigate maximum likelihood-expectation maximization (ML-EM) methods to reconstruct simultaneously acquired parallel and cone beam (P&CB) SPECT data. Simultaneous P&CB acquisition can be performed with commercially available triple camera systems by using two cone-beam collimators and a single parallel-hole collimator. The loss in overall sensitivity (relative to the use of three CB collimators) is about 15 to 20%. We have developed three methods to combine P&CB data using modified ML-EM algorithms. The first method consists of using both data sets to reconstruct a single intermediate image after each iteration using the ML-EM algorithm. The other two iterative algorithms combine intermediate parallel beam (PB) and CB source estimates to enhance image quality. For these methods, a PB estimate and a CB estimate are obtained for the first iteration. The second method consists of summing the PB and CB estimates for each subsequent iteration to obtain new PB and CB estimates. The third method is similar to the second method, with the exception that the new PB estimate simply is set equal to the PB estimate after each iteration. The combined source estimate is used in each subsequent iteration step of the EM algorithm. These algorithms are evaluated using projection data simulated using a Monte Carlo SPECT model. The P&CB SPECT images demonstrate marked improvement as compared with the CB-only reconstruction, particularly when the projections are truncated.

Algorithms↗
PubMed2006

A multivariate model for ordinal trait analysis.

Many economically important characteristics of agricultural crops are measured as ordinal traits. Statistical analysis of the genetic basis of ordinal traits appears to be quite different from regular quantitative traits. The generalized linear model methodology implemented via the Newton-Raphson algorithm offers improved efficiency in the analysis of such data, but does not take full advantage of the extensive theory developed in the linear model arena. Instead, we develop a multivariate model for ordinal trait analysis and implement an EM algorithm for parameter estimation. We also propose a method for calculating the variance-covariance matrix of the estimated parameters. The EM equations turn out to be extremely similar to formulae seen in standard linear model analysis. Computer simulations are performed to validate the EM algorithm. A real data set is analyzed to demonstrate the application of the method. The advantages of the EM algorithm over other methods are addressed. Application of the method to QTL mapping for ordinal traits is demonstrated using a simulated baclcross (BC) population.

Algorithms↗
PubMed1996

Testing for linkage disequilibrium in genotypic data using the Expectation-Maximization algorithm.

We generalize an approach suggested by Hill (Heredity, 33, 229-239, 1974) for testing for significant association among alleles at two loci when only genotype and not haplotype frequencies are available. The principle is to use the Expectation-Maximization (EM) algorithm to resolve double heterozygotes into haplotypes and then apply a likelihood ratio test in order to determine whether the resolutions of haplotypes are significantly nonrandom, which is equivalent to testing whether there is statistically significant linkage disequilibrium between loci. The EM algorithm in this case relies on the assumption that genotype frequencies at each locus are in Hardy-Weinberg proportions. This method can accommodate X-linked loci and samples from haplodiploid species. We use three methods for testing significance of the likelihood ratio: the empirical distribution in a large number of randomized data sets, the X2 approximation for the distribution of likelihood ratios, and the Z2 test. The performance of each method is evaluated by applying it to simulated data sets and comparing the tail probability with the tail probability from Fisher's exact test applied to the actual haplotype data. For realistic sample sizes (50-150 individuals) all three methods perform well with two or three alleles per locus, but only the empirical distribution is adequate when there are five to eight alleles per locus, as is typical of hypervariable loci such as microsatellites. The method is applied to a data set of 32 microsatellite loci in a Finnish population and the results confirm the theoretical predictions. We conclude that with highly polymorphic loci, the EM algorithm does lead to a useful test for linkage disequilibrium, but that it is necessary to find the empirical distribution of likelihood ratios in order to perform a test of significance correctly.

Algorithms↗
PubMed2003

The EIM algorithm in the joint segregation analysis of quantitative traits.

In this article, a new algorithm for obtaining the maximum likelihood estimators (MLEs) of parameters in the joint segregation analysis (JSA) of multiple generations of P1, F1, P2, F2 and F2:3 (MG5) for quantitative traits was set up. Firstly, owing to the fact that the component variance of the heterogeneous genotype in F2:3 included both the first-order genetic parameters (denoted by the means of distributions) and the second-order parameters, a simple closed form for the MLEs of the means of component distributions did not exist while the expectation and maximization (EM) algorithm was used. To simplify the estimation of parameters, the first partial derivative of the above variance on the mean in the sample log-likelihood function was omitted. However, this would be remedied by the iterated method. Then, variances of component distributions for segregating populations were partitioned into major-gene, polygenic and environmental variances so that the generally iterated formulae for estimating the means as well as polygenic and environmental variances of component distributions in the maximization step (M-step) of the EM algorithm were obtained. Therefore, the EM algorithm for estimating parameters in the JSA model for the MG5 was simplified. This is called the expectation and iterated maximization (EIM) algorithm. Finally, an example of the inheritance of the resistance of soybean to beanfly showed that the results of mixed inheritance analysis in this paper coincided with those in both Wang & Gai (2001) and Wei et al. (1989), so the EIM algorithm was appropriate.

Algorithms↗
PubMed2001

Mosaic model for sensorimotor learning and control.

Humans demonstrate a remarkable ability to generate accurate and appropriate motor behavior under many different and often uncertain environmental conditions. We previously proposed a new modular architecture, the modular selection and identification for control (MOSAIC) model, for motor learning and control based on multiple pairs of forward (predictor) and inverse (controller) models. The architecture simultaneously learns the multiple inverse models necessary for control as well as how to select the set of inverse models appropriate for a given environment. It combines both feedforward and feedback sensorimotor information so that the controllers can be selected both prior to movement and subsequently during movement. This article extends and evaluates the MOSAIC architecture in the following respects. The learning in the architecture was implemented by both the original gradient-descent method and the expectation-maximization (EM) algorithm. Unlike gradient descent, the newly derived EM algorithm is robust to the initial starting conditions and learning parameters. Second, simulations of an object manipulation task prove that the architecture can learn to manipulate multiple objects and switch between them appropriately. Moreover, after learning, the model shows generalization to novel objects whose dynamics lie within the polyhedra of already learned dynamics. Finally, when each of the dynamics is associated with a particular object shape, the model is able to select the appropriate controller before movement execution. When presented with a novel shape-dynamic pairing, inappropriate activation of modules is observed followed by on-line correction.

Brain↗
PubMed1998

Evaluation of a cross-validation stopping rule in MLE SPECT reconstruction.

One of the problems in the routine use of the maximum-likelihood estimator method-expectation maximization (MLE-EM) algorithm is to decide when the iterative process should be stopped. We studied a cross-validation stopping rule to assess its usefulness in SPECT. We tested this stopping rule criterion in the MLE-EM algorithm without acceleration as well as in two accelerating algorithms, the successive substitutions algorithm (SSA) and the additive algorithm (AA). Different values of an acceleration factor were tested in SSA and AA. Our results from numerical and physical phantoms show that the stopping rule based on the cross-validation ratio (CVR) takes into account the similarity of the reconstructed image to the ideal image, noise and the contrast of the image. CVR yields reconstructed images with balanced values of the figures of merit (FOM) employed to assess the image quality. The CVR criterion can be used in the original MLE-EM algorithm as well as in SSA and AA. The reconstructed images obtained with SSA and AA showed FOM values that were very similar. These results were justified by considering AA to be an approximate form of SSA. The range of validity for the acceleration factor in SSA and AA was found to be [1, 2]. In this range, an inverse function connects the acceleration factor to the number of iterations needed to attain prefixed values of FOMs.

Algorithms↗
PubMed1998

Block-iterative techniques for fast 4D reconstruction using a priori motion models in gated cardiac SPECT.

We introduce a fast block-iterative maximum a posteriori (MAP) reconstruction algorithm and apply it to four-dimensional reconstruction of gated SPECT perfusion studies. The new algorithm, called RBI-MAP, is based on the rescaled block iterative EM (RBI-EM) algorithm. We develop RBI-MAP based on similarities between the RBI-EM, ML-EM and MAP-EM algorithms. RBI-MAP requires far fewer iterations than MAP-EM, and so should result in acceleration similar to that obtained from using RBI-EM or OS-EM as opposed to ML-EM. When complex four-dimensional clique structures are used in the prior, however, evaluation of the smoothing prior dominates the processing time. We show that a simple scheme for updating the prior term in the heart region only for RBI-MAP results in savings in processing time of a factor of six over MAP-EM. The RBI-MAP algorithm incorporating 3D collimator-detector response compensation is demonstrated on a simulated 99mTc gated perfusion study. Results of RBI-MAP are compared with RBI-EM followed by a 4D linear filter. For the simulated study, we find that RBI-MAP provides consistently higher defect contrast for a given degree of noise smoothing than does filtered RBI-EM. This is an indication that RBI-MAP smoothing does less to degrade resolution gained from 3D detector response compensation than does a linear filter. We conclude that RBI-MAP can provide smooth four-dimensional reconstructions with good visualization of heart structures in clinically realistic processing times.

Algorithms↗