Aspect-based sentiment analysis is the most important research topic conducted to extract and categorize aspect-terms from online reviews. Recent efforts have shown that topic modelling is vigorously used for this task. In this paper, we integrated word embedding into collapsed Gibbs sampling in Latent Dirichlet Allocation (LDA). Specifically, the conditional distribution in the topic model is improved using the word embedding model that was trained against (customer review) training dataset. Semantic similarity (cosine measure) was leveraged to distribute the aspect-terms to their related aspect-category cognitively. The experiment was conducted to extract and categorize the aspect terms from SemEval 2014 dataset.
In this paper, a Bayesian analysis is made to estimate the Reliability of two stress-strength model systems. First: the reliability of a one component strengths X under stress Y. Second, reliability of one component strength under three stresses. Where X and Y are independent generalized exponential-Poison random variables with parameters (α,λ,θ) and (β,λ,θ) . The analysis is concerned with and based on doubly type II censored samples using gamma prior under four different loss functions, namely quadratic loss function, weighted loss functions, linear and non-linear exponential loss function. The estimators are compared by mean squared error criteria due to a simulation study. We also find that the mean square error is
... Show MoreThis paper concerns with deriving and estimating the reliability of the multicomponent system in stress-strength model R(s,k), when the stress and strength are identical independent distribution (iid), follows two parameters Exponentiated Pareto Distribution(EPD) with the unknown shape and known scale parameters. Shrinkage estimation method including Maximum likelihood estimator (MLE), has been considered. Comparisons among the proposed estimators were made depending on simulation based on mean squared error (MSE) criteria.
This Book is the second edition that intended to be textbook studied for undergraduate/ postgraduate course in mathematical statistics. In order to achieve the goals of the book, it is divided into the following chapters. Chapter One introduces events and probability review. Chapter Two devotes to random variables in their two types: discrete and continuous with definitions of probability mass function, probability density function and cumulative distribution function as well. Chapter Three discusses mathematical expectation with its special types such as: moments, moment generating function and other related topics. Chapter Four deals with some special discrete distributions: (Discrete Uniform, Bernoulli, Binomial, Poisson, Geometric, Neg
... Show MoreIn this paper, we investigate the behavior of the bayes estimators, for the scale parameter of the Gompertz distribution under two different loss functions such as, the squared error loss function, the exponential loss function (proposed), based different double prior distributions represented as erlang with inverse levy prior, erlang with non-informative prior, inverse levy with non-informative prior and erlang with chi-square prior.
The simulation method was fulfilled to obtain the results, including the estimated values and the mean square error (MSE) for the scale parameter of the Gompertz distribution, for different cases for the scale parameter of the Gompertz distr
... Show MoreThis paper is concerned with preliminary test double stage shrinkage estimators to estimate the variance (s2) of normal distribution when a prior estimate of the actual value (s2) is a available when the mean is unknown , using specifying shrinkage weight factors y(×) in addition to pre-test region (R).
Expressions for the Bias, Mean squared error [MSE (×)], Relative Efficiency [R.EFF (×)], Expected sample size [E(n/s2)] and percentage of overall sample saved of proposed estimator were derived. Numerical results (using MathCAD program) and conclusions are drawn about selection of different constants including in the me
... Show MoreLet R be a commutative ring with unity and an R-submodule N is called semimaximal if and only if
the sufficient conditions of F-submodules to be semimaximal .Also the concepts of (simple , semisimple) F- submodules and quotient F- modules are introduced and given some properties .
Background: Alcohol remains the single most significant cause of liver disease throughout the Western world, responsible for between 40 and 80% of cases of cirrhosis in different countries. Many of the factors underlying the development of alcoholic liver injury remain unknown, and significant questions remain about the value of even very basic therapeutic strategies.
Patients and Methods: In a cross sectional study, 113 alcoholic patients with evidence of liver disease in the absence of other significant etiology attending the Gastoenterorology and Hepatology Teaching Hospital between December 2001 and December 2003 were studied for the hematological and biochemical spectrum of alcoholic liver disease in