Search results for "Expectation–maximization algorithm"
showing 10 items of 25 documents
On properties of the iterative maximum likelihood reconstruction method
1989
In this paper, we continue our investigations6 on the iterative maximum likelihood reconstruction method applied to a special class of integral equations of the first kind, where one of the essential assumptions is the positivity of the kernel and the given right-hand side. Equations of this type often occur in connection with the determination of density functions from measured data. There are certain relations between the directed Kullback–Leibler divergence and the iterative maximum likelihood reconstruction method some of which were already observed by other authors. Using these relations, further properties of the iterative scheme are shown and, in particular, a new short and elementar…
Multiple imputation of rainfall missing data in the Iberian Mediterranean context
2017
Abstract Given the increasing need for complete rainfall data networks, in recent years have been proposed diverse methods for filling gaps in observed precipitation series, progressively more advanced that traditional approaches to overcome the problem. The present study has consisted in validate 10 methods (6 linear, 2 non-linear and 2 hybrid) that allow multiple imputation, i.e., fill at the same time missing data of multiple incomplete series in a dense network of neighboring stations. These were applied for daily and monthly rainfall in two sectors in the Jucar River Basin Authority (east Iberian Peninsula), which is characterized by a high spatial irregularity and difficulty of rainfa…
Graph Topology Learning and Signal Recovery Via Bayesian Inference
2019
The estimation of a meaningful affinity graph has become a crucial task for representation of data, since the underlying structure is not readily available in many applications. In this paper, a topology inference framework, called Bayesian Topology Learning, is proposed to estimate the underlying graph topology from a given set of noisy measurements of signals. It is assumed that the graph signals are generated from Gaussian Markov Random Field processes. First, using a factor analysis model, the noisy measured data is represented in a latent space and its posterior probability density function is found. Thereafter, by utilizing the minimum mean square error estimator and the Expectation M…
The EM imaging reconstruction method in γ-ray astronomy
1998
Abstract The simpler imaging reconstruction methods used for γ-ray coded mask telescopes are based on correlation methods, very fast and simple-to-use but with limitations in the reconstructed image. To improve these results, other reconstruction methods have been developed, such as the maximum entropy methods or the Iterative Removal Of Sources (IROS). However, such kind of methods are slower and can be impracticable for very complex telescopes. In this paper we present an alternative image reconstruction method, based on an iterative maximum likelihood algorithm called the EM algorithm, easy to implement and that can be successfully used for not very complex coded mask systems, as is the …
Modelling of Magnetic Resonance Spectra Using Mixtures for Binned and Truncated Data
2007
Magnetic Resonance Spectroscopy (MRS) provides the biochemical composition of a tissue under study. This information is useful for the in-vivo diagnosis of brain tumours. Prior knowledge of the relative position of the organic compound contributions in the MRS suggests the development of a probabilistic mixture model and its EM-based Maximum Likelihood Estimation for binned and truncated data. Experiments for characterizing and classifying Short Time Echo (STE) spectra from brain tumours are reported.
A probabilistic framework for automatic prostate segmentation with a statistical model of shape and appearance
2011
International audience; Prostate volume estimation from segmented prostate contours in Trans Rectal Ultrasound (TRUS) images aids in diagnosis and treatment of prostate diseases, including prostate cancer. However, accurate, computationally efficient and automatic segmentation of the prostate in TRUS images is a challenging task owing to low Signal-To-Noise-Ratio (SNR), speckle noise, micro-calcifications and heterogeneous intensity distribution inside the prostate region. In this paper, we propose a probabilistic framework for propagation of a parametric model derived from Principal Component Analysis (PCA) of prior shape and posterior probability values to achieve the prostate segmentatio…
Morphological exponential entropy driven-HUM.
2006
This paper presents an improvement to the Ex- ponential Entropy Driven - Homomorphic Unsharp Masking (E 2 D − HUM ) algorithm devoted to illumination artifact sup- pression on Magnetic Resonance Images. E 2 D−HUM requires a segmentation step to remove dark regions in the foreground whose intensity is comparable with background, because strong edges produce streak artifacts on the tissues. This new version of the algorithm keeps the same good properties of E 2 D − HUM without a segmentation phase, whose parameters should be chosen in relation to the image. I. INTRODUCTION Most of the studies on illumination correction found in literature are oriented to brain (18) magnetic resonance images (…
Spectrum cartography using adaptive radial basis functions: Experimental validation
2017
In this paper, we experimentally validate the functionality of a developed algorithm for spectrum cartography using adaptive Gaussian radial basis functions (RBF). The RBF are strategically centered around representative centroid locations in a machine learning context. We assume no prior knowledge about neither the power spectral densities (PSD) of the transmitters nor their locations. Instead, the received signal power at each location is estimated as a linear combination of different RBFs. The weights of the RBFs, their Gaussian decaying parameters and locations are jointly optimized using expectation maximization with a least squares loss function and a quadratic regularizer. The perfor…
Spatio‐temporal classification in point patterns under the presence of clutter
2019
We consider the problem of detection of features in the presence of clutter for spatio-temporal point patterns. In previous studies, related to the spatial context, Kth nearest-neighbor distances to classify points between clutter and features. In particular, a mixture of distributions whose parameters were estimated using an expectation-maximization algorithm. This paper extends this methodology to the spatio-temporal context by considering the properties of the spatio-temporal Kth nearest-neighbor distances. For this purpose, we make use of a couple of spatio-temporal distances, which are based on the Euclidean and the maximum norms. We show close forms for the probability distributions o…
Forecasting time series with missing data using Holt's model
2009
This paper deals with the prediction of time series with missing data using an alternative formulation for Holt's model with additive errors. This formulation simplifies both the calculus of maximum likelihood estimators of all the unknowns in the model and the calculus of point forecasts. In the presence of missing data, the EM algorithm is used to obtain maximum likelihood estimates and point forecasts. Based on this application we propose a leave-one-out algorithm for the data transformation selection problem which allows us to analyse Holt's model with multiplicative errors. Some numerical results show the performance of these procedures for obtaining robust forecasts.