We propose statistical abduction as a rstorder logical framework for representing and learning probabilistic knowledge. It combines logical abduction with a parameterized distribution over abducibles. We show that probability computation, a Viterbilike algorithm and EM learning for statistical abduction achieve the same eÆciency as specilzed algorithms for HMMs (hidden Markov models), PCFGs (pr...

We present a general framework for unsupervised and semi-supervised learning containing a graded spectrum of Expectation Maximization (EM) algorithms. We call our framework Unified Expectation Maximization (UEM.) UEM allows us to tune the entropy of the inferred posterior distribution during the E-step to impact the quality of learning. Furthermore, UEM covers existing algorithms like standard ...

Based on the electromagnetism-like algorithm (EM), we propose a novel hybrid learning algorithms which is the improved EM algorithm with genetic algorithm technique (IEMGA) for recurrent fuzzy neural system design. IEMGA are composed of initialization, local search, total force calculation, movement, and evaluation. They are hybridization of EM and GA. EM algorithm is a population-based meta-he...

Most of the researchers in the application areas usually use the EM algorithm to *nd estimators of the normal mixture distribution with unknown component speci*c variances without knowing much about the properties of the estimators. It is unclear for which situations the EM algorithm provides “good” estimators, good in the sense of statistical properties like consistency, bias, or mean square e...

texture image analysis is one of the most important working realms of imageprocessing in medical sciences and industry. up to present, different approacheshave been proposed for segmentation of texture images. in this paper, we offeredunsupervised texture image segmentation based on markov random field (mrf)model. first, we used gabor filter with different parameters’ (frequency,orientation) va...

It’s very important for us to understand the data structure before doing the data analysis. However, most of the time, there may exist of a lot of missing values or incomplete information in the data subject to the analysis. For example, survival time data always have some missing values because of death or job transfer. These kinds of data are called censored data. Since these data might obtai...

New iterative algorithms are presented for Maximum Likelihood (ML) and Regularized Maximum Likelihood (MAP) reconstruction in Transmission Tomography (CT). The algorithms are natural extensions to CT of RAMLA, a well known method for ML reconstruction in Emission Computed Tomography (ECT). We show that the new algorithm for ML solutions produces similar, or even better results than EM-like algo...

‎in this paper, we study a flow shop batch processing machines scheduling problem. the fuzzy due dates are considered to make the problem more close to the reality. the objective function is taken as the weighted sum of fuzzy earliness and fuzzy tardiness. in order to tackle the given problem, we propose a hybrid electromagnetism-like (em) algorithm, in which the em is hybridized with a diversi...

In this study, we would like to present brain tumor detection methods, based on the conventional K-means technique, Expectation Maximization (EM) algorithm and a new Spatial Fuzzy-technique analysis of brain MR images. Though, the Kmeans and EM algorithm were already used in Brain MR image segmentation, as well as image segmentation in general, it fails to utilize the strong spatial correlation...

In this paper, we present on a periodic job shop scheduling problem (PJSSP) based on the periodic event scheduling problem (PESP), which is different from cyclic scheduling. The PESP schedules a number of recurring events, such that each pair of events fulfills certain constraints during a given time period. To solve such a hard PJSS problem, we propose a hybrid algorithm, namely EM–SA, which i...