Showing papers on "Maximum a posteriori estimation published in 1980"
284 citations
TL;DR: In this article, modifications of local maximum likelihood estimators and modifications of moment estimators for parameters of the three-parameter lognormal distribution are discussed, where first, second and third order statistics are employed to facilitate estimation of the distribution lower limit.
Abstract: This article is primarily concerned with modifications of local maximum likelihood estimators and modifications of moment estimators for parameters of the three-parameter lognormal distribution. First, second, and third order statistics are employed to facilitate estimation of the distribution lower limit. Computational procedures are developed to simplify the practical application of these estimators. Results of a simulation study provide insight into the sampling behavior of the estimates.
106 citations
TL;DR: In this paper, a Mann-Whitney type statistic is used to estimate a change-point when a change, at an unknown point in a sequence of random variables, has taken place.
Abstract: A Mann-Whitney type statistic is used to estimate a change-point when a change, at an unknown point in a sequence of random variables, has taken place. This estimate is compared, using Monte Carlo techniques, with the normal theory maximum likelihood estimate, when a location change has occurred, for different underlying distributions ranging from the normal to the long tailed “normal over uniform” distribution. The distribution of the Mann-Whitney type estimate remains fairly constant over the various distributions. Two generalisations of the statistic are considered and investigated.
45 citations
TL;DR: It is shown that the ME solution is a member of the set of MAP solutions defined by a set of a priori probability densities.
Abstract: For the linear imaging model g = Hf + n, the maximum a posteriori (MAP) restoration method is compared to the maximum entropy (ME) method defined by maximizing f_{T} \ln f subject to \parallelg - Hf\parallel^{2} = \paralleln\parallel^{2} . It is shown that the ME solution is a member of the set of MAP solutions defined by a set of a priori probability densities. The numerical methods developed for MAP restoration can be applied to ME restoration. The importance of the a priori probability distribution for the MAP restoration is demonstrated. Examples of ME restoration with the new method are shown and compared to previous results.
41 citations
TL;DR: In this paper, the bias correction to the maximum likelihood extimates of the parameters for logistic discrimination is examined under mixture and separate sampling schemes, and the effect of this bias correction on the sampling properties of the misclassification rates and of the estimated posterior probabilities is discussed.
Abstract: The bias correction to the maximum likelihood extimates of the parameters for logistic discrimination is examined under mixture and separate sampling schemes. An existing adjustment developed under mixture sampling and based on higher derivatives of the log likelihood is modified slightly for use under separate sampling. The effect of this bias correction on the sampling properties of the misclassification rates and of the estimated posterior probabilities is discussed.
33 citations
TL;DR: In this article, maximum a posteriori (MAP) techniques are applied to the problem of estimating the unknown parameters of a frequency modulated narrowband signal in additive white Gaussian noise.
Abstract: Maximum a posteriori (MAP) techniques are applied to the problem of estimating the unknown parameters of a frequency modulated narrow‐band signal in additive white Gaussian noise. Signals are assumed to have unknown amplitude, initial phase, and frequency versus time history, although a priori information concerning the frequency trajectory may be available. In addition, it is assumed that the signal frequency trajectory can be adequately modeled by a continuous piecewise linear function of time. It is shown that the MAP estimator maximizes a linear combination of (1) coherent match of signal induced outputs to received observations, and (2) a term dependent only upon a priori knowledge. This second term reduces to a measure of trajectory smoothness in the special case of a Gauss–Markov a priori frequency model. Lastly, a MAP detection scheme is suggested for cases in which the presence of such signals is an issue.
22 citations
17 citations
TL;DR: In this paper, almost sure convergence of the maximum likelihood and the maximum a posteriori probability estimates of unknown parameters of continuous-time stochastic dynamical linear time-invariant systems is investigated.
Abstract: Almost-sure convergence of the maximum likelihood and the maximum a posteriori probability estimates of unknown parameters of continuous-time stochastic dynamical linear time-invariant systems is investigated. The unknown parameter set is assumed to be finite. The situation where the ture parameter does not belong to the unknown parameter set is considered, as well as the situation where the true model is included in the unknown parameter set.
12 citations
01 Dec 1980
TL;DR: A general architecture is presented to solve the multisensor, multitarget recognition and tracking problem and the treatment of multispectral noncommensurate attributes via their target classification capability is derived.
Abstract: A general architecture is presented to solve the multisensor, multitarget recognition and tracking problem. Candidate maximum a posteriori objective functionals are presented for kinematic and attribute report correlation. The treatment of multispectral noncommensurate attributes via their target classification capability is derived.
11 citations
TL;DR: In this article, a simulation experiment comparing the accuracy and precision of three alternate estimation techniques for the parameters of the STARMA model is presented. But the accuracy of these three estimation procedures for simulated series of various lengths, and the appropriateness of the three procedures as a function of the observed series is discussed.
Abstract: A simulation experiment compares the accuracy and precision of three alternate estimation techniques for the parameters of the STARMA model. Maximum likelihood estimation, in most ways the "best" estimation procedure, involves a large amount of computational effort so that two approximate techniques, exact least squares and conditional maximum likelihood, are often proposed for series of moderate lengths. This simulation experiment compares the accuracy of these three estimation procedures for simulated series of various lengths, and discusses the appropriateness of the three procedures as a function of the length of the observed series.
TL;DR: In this paper, the authors consider truncated samples taken from an infinite population with a fixed number of observations recorded, and propose a computing scheme for the exponential distribution on the basis of asymptotu,operties.
Abstract: Consider truncated samples taken from an infinite population with a fixed uumber n of observations recorded A randon number X of items must be sampled in order to observe after trunca-tion adified maximun liicelihooa estimators of X (assumed un-'mown) and of the population parameters are develo~cd, and a computing scheme is given for the exponential distribution On the basis of asymptotu ,operties, some estimators are singled out and compared with the usual maximum likelihood estimators
01 Dec 1980
TL;DR: In this article, a truncated maximum likelihood estimation scheme is proposed for consistent estimation of the transition probabilities in a linear discrete-time system with interrupted observations, where the interrupted observation mechanism is expressed in terms of a two-state Markov chain.
Abstract: A linear discrete-time system with interrupted observations is considered. The interrupted observation mechanism is expressed in terms of a two-state Markov chain. The transition probability matrix of the Markov chain is unknown and is assumed to belong to a compact set. A novel scheme, called truncated maximum likelihood estimation, is proposed for consistent estimation of the transition probabilities. Conditions for weak consistency are investigated.
01 Dec 1980
TL;DR: In this paper, a robust filter and smoother is applied to tracking data to obtain improved estimation performance in the presence of outliers, and the improvement in estimation performance is evaluated by Monte Carlo using simulated tracking data.
Abstract: : Robust methods provide a fresh approach to the treatment of outliers in filtering and smoothing applications. In deriving the filter and smoother equations via the conditional mean formulation or maximum a posteriori formulation the measurement noise probability density is replaced by a pseudo density which is Gaussian mixture with very heavy tails. The resulting robust filter and smoother are applied to tracking data to obtain improved estimation performance in the presence of outliers. The improvement in estimation performance is evaluated by Monte Carlo using simulated tracking data. The Monte Carlo results indicate the improvement in performance to be somewhat greater than the improvement obtained when using robust filters and smoothers derived from M-estimates. (Author)
01 Dec 1980
TL;DR: In this paper, the asymptotic behavior of Bayes optimal adaptive state estimation schemes (also called the partitioned adaptive estimation algorithms) for continuous-time linear dynamic Gauss-Markov systems with unknown parameters is investigated.
Abstract: The asymptotic behavior of Bayes optimal adaptive state estimation schemes (also called the partitioned adaptive estimation algorithms) for continuous-time linear dynamic Gauss-Markov systems with unknown parameters is investigated. The unknown system parameters are assumed to belong to a finite set. The results are developed through weak consistency of the maximum likelihood and the maximum a posteriori probability estimates of the unknown parameters.
01 Feb 1980
TL;DR: In this paper, the reliability function for the inverse Gaussian distribution is discussed and Bayes estimators are obtained with Jeffreys' noninformative prior and with the natural conjugate prior for the scale parameter.
Abstract: : Bayes estimation of the reliability function for the inverse Gaussian distribution is discussed. For the case that the mean lifetime is known, Bayes estimators are obtained with Jeffreys' noninformative prior and with the natural conjugate prior for the scale parameter. In the case that both parameters are unknown, an estimator of reliability is suggested which is based on the Bayes estimator obtained for the case that the mean lifetime is known. This estimator is not Bayes but compares favorably with the maximum likelihood and minimum variance unbiased estimators as indicated by computer simulations. (Author)
TL;DR: By adopting the Bayesian approach, an alternative procedure for the estimation of linkage was proposed in this article, and the proposed method was then compared with the classical maximum likelihood method, especially in the repulsion phase.
Abstract: By adopting the Bayesian approach, this paper proposes an alternative procedure for the estimation of linkage.Through simulation studies, the proposed method was then compared with the classical maximum likelihood method. The numerical results indicate clearly that in most of the cases, the Bayesian method is better than the maximum likelihood method, especially in the repulsion phase.