Earticle

Previous studies present the mixed results on online reputation mechanism. In this study, we have found that an approach based on Bayesian statistics can explain most results of previous studies which are conflicting with each others. With this model, we explain why negative ratings have more significant marginal impacts on sellers' reputation than positive ones do. Furthermore, we even show why the feedbacks with a few negative ratings may increase the value of the item and final prices by confirming buyers' prior beliefs on the sellers' reputation much more than those without negative ratings. Also, we explain why there are not many negative ratings. Even though some studies suggest this because of generosity of users, our model shows that the reason is that the existence of FS itself prevents bad sellers from participating to the market as a signal itself. Even further, we show how this extreme tendency of positive ratings gets even stronger as markets evolve. Finally, to validate our analytical results, we examine the previous studies and see what factors effect the outcomes of their analyses.

In this paper multiple senses of some Korean ambiguous words are discriminated on the basis of Bayesian inference which utilizes the conditional probability widely accepted in mathematics. A POS tagged 8.1 million words Korean corpus was used as the resource of the linguistic informations for disambiguation. As a result of disambiguational experiment on the 13 words(9 nouns and 4 verbs) by computational programming of the algorithm based on the Bayesian inference, the whole precision accomplished 81.5%(25981/31874), with 83.5%(12546/15030) for nouns and 79.8%(13435/16844) for verbs respectively. In the course of the experiment some parametric variations were engaged to reveal the optimistic condition for this methodological process. The focus was set on the effect of the variation of the smoothing values from 0.9 to 0.0001 which is substituted for the value 0 of the co-occurrence frequency of a word in the context, and to the contrary of general expectations, smoothing value 0.1 resulted in the topmost precision. In addition to the machine process and its promising result, the way how the individual words of the sentences in the corpus are to be treated under the Bayesian inference is exemplified in this paper in detail, thus clarifying the methodological understanding.

Cloud Computing plays an important role in academic and industry. Business process are composed and executed in the distributed loosely coupled environment. Business processes are implemented as the composite service which involves many services and these services connected by different workflow patterns. Numerous functionally similar services are available in the cloud. As it is an on-demand service, cloud service consumer needs some Quality of Service factors in services. Trust plays an import role in service selection. Selecting the trust worthy service for service composition with the cost and other constraints is a tedious task. In this paper, service trust estimation method based on Beta distribution is proposed. Trust based Cloud service selection framework for service composition is proposed and it is proved that the proposed framework provides better performance as compared to the well-known service composition methods based on the execution time and optimality value.

Variational Bayesian (VB) inference is the latest iterative method for prediction of data in machine learning. It provides the solution for intractable integration in Bayesian methodology. In this paper, a simple VB linear regression is applied for prediction of the damaged pixels in an image. Bayesian linear regression model is used for prediction of the pixels. For this neighbor pixels are used as training data to generate the parameters of the prediction function. Now using this prediction function, damaged pixels are predicted and incorporated into the image. Proposed method is linear while image is a non-linear object, generally. Hence, for linearity, a small image window size is used to avoid the nonlinearities in image.

Taking account of the influence of uncertainty, incomplete issues and common cause failure (CCF) to system reliability, a method for reliability analysis based on Dempster-Shafer (D-S) evidence inference and Bayesian network (BN) is proposed by using the advantage of uncertainty reasoning and figurative expression of BN. Firstly, the method to map fault tree into BN with epistemic uncertainty is presented, and the probability of top event (TE) is calculated by belief measure and plausibility measure of evidence. Moreover, by using BN, some useful information can be obtained, through which the weak parts of the system can be identified. The results demonstrate that the proposed method makes a more rational evaluation result, and it is quite valid to deal with uncertainty information in reliability analysis, which provide a reasonable guidance for system fault diagnosis and maintenance and improve the system reliability.

마케팅 분야에서는 제품품질, 고객만족, 고객추천을 바탕으로 구매행동과의 영향 유무 및 상관관계를 통계적 Regression 방법으로 가설 검증하는 것을 주요한 연구 대상으로 하고 있다. 또한 최근에는 ASCI와 같은 고객만족지수 혹은 라이켈트의 NPS와 같은 고객추천지수를 바탕으로 실제 기업성과와 연관되는 시장 지분에 어떠한 영향을 미치는 지에 대한 통계적 분석 연구도 활발히 이루어지고 있다. 본 연구에서는 실제 고객이 매장을 방문하여, 과거 고객카드에 명품을 구매하던 구매하지 않던 간에 만족/불만족을 표시한 체인 및 고객 추천의향을 검토하여 Hidden Markov Model을 이용한 고객의 최상의 구매패턴을 분석하는 확률적 기법에 대하여 연구하는 것을 목적으로 하고 있다. 이를 바탕으로 고객만족 -> 고객추천의향 -> 고객추천행동->구매 및 재구매 체인에 대응하는 실제 소비자의 구매패턴을 고객만족과 NPS(순추천지수) 및 여러 수리통계적 이론-Hidden Markov Model, Bayesian Inference, Maximum Likelihood Estimation을 이용하여 확률적 추적 메카니즘을 구현하는 것을 목표로 한다. 제시된 목표는 인공지능을 구현하는 이론과 알고리듬을 사용하여 달성되었기에 이론적 추적 메카니즘을 여러 인공지능망 -DNN, CNN, GAN등을 사용하여 기업에서 사용할 수 있는 고객의 구매패턴 앱으로 발전시키는 것을 후속연구에서 기대한다.

The purpose of this study is to specify a probabilistic tracking mechanism for customer luxury purchase implemented by hidden Markov model, Bayesian inference, customer satisfaction and net promoter score. In this paper, we have designed a probabilistic model based on customer's actual data containing purchase or non-purchase states by tracking the SPC chain : customer satisfaction -> customer referral -> purchase/non-purchase. By applying hidden Markov model and Viterbi algorithm to marketing theory, we have developed the statistical model related to probability theories and have found the best purchase pattern scenario from customer's purchase records.

We propose a Bayesian approach to cumulative logistic regression model for the ordinal response based on the orthogonal principal components via singular value decomposition considering the multicollinearity among predictors. The advantage of the suggested method is considering dimension reduction and parameter estimation simultaneously. To evaluate the performance of the proposed model we conduct a simulation study with considering a high-dimensional and highly correlated explanatory matrix. Also, we fit the suggested method to a real data concerning sprout- and scab-damaged kernels of wheat and compare it to EM based proportional-odds logistic regression model. Compared to EM based methods, we argue that the proposed model works better for the highly correlated high-dimensional data with providing parameter estimates and provides good predictions.

A stochastic Gompertz diffusion model for tumor growth is a topic of active interest as cancer is a leading cause of death in Korea. The direct maximum likelihood estimation of stochastic differential equations would be possible based on the continuous path likelihood on condition that a continuous sample path of the process is recorded over the interval. This likelihood is useful in providing a basis for the so-called continuous record or infill likelihood function and infill asymptotic. In practice, we do not have fully continuous data except a few special cases. As a result, the exact ML method is not applicable. In this paper we proposed a method of parameter estimation of stochastic Gompertz differential equation via Markov chain Monte Carlo methods that is applicable for several data structures. We compared a Markov transition data structure with a data structure that have an initial point.

It was recognized by some researchers that the disturbance variance in a censored regression model is frequently underestimated by the maximum likelihood method. This underestimation has implications for the estimation of marginal effects and asymptotic standard errors. For instance, the actual coverage probability of the confidence interval based on a maximum likelihood estimate can be significantly smaller than the nominal confidence level; consequently, a Bayesian estimation is considered to overcome this difficulty. The behaviors of the maximum likelihood and Bayesian estimators of disturbance variance are examined in a fixed effects panel regression model with a limited dependent variable, which is known to have the incidental parameter problem. Behavior under random effect assumption is also investigated.

The paper considers empirical Bayes (EB) and hierarchical Bayes (HB) predictors of the finite population mean under a linear regression model with measurement errors We discuss how to calculate the mean squared prediction errors of the EB predictors using jackknife methods and the posterior standard deviations of the HB predictors based on the Markov Chain Monte Carlo methods. A simulation study is provided to illustrate the results of the preceding sections and compare the performances of the proposed procedures.

This paper consider trinomial group testing concerned with classification of N given units into one of k disjoint categories. In this paper, we propose Bayesian inference for estimating individual category proportions using the trinomial group testing model proposed by Bar-Lev et al. (2005). We compared a relative efficience (RE) based on the mean squared error (MSE) of MLE and Bayes estimators with various prior information. The impact of different prior specifications on the estimates is also investigated using selected prior distribution. The impact of different priors on the Bayes estimates is modest when the sample size and group size we large.

In this paper, we consider Bayesian approach to the Fieller-Creasy problem using noninformative priors. Specifically we extend the results of Yin and Ghosh (2000) to the unbalanced case. We develop some noninformative priors such as the first and second order matching priors and reference priors. Also we prove the posterior propriety under the derived noninformative priors. We compare these priors in light of how accurately the coverage probabilities of Bayesian credible intervals match the corresponding frequentist coverage probabilities.

Commercial banks and other related areas have developed internal models to better quantify their financial risks. Since an appropriate credit risk model plays a very important role in the risk management at financial institutions, it needs more accurate model which forecasts the credit losses, and statistical inference on that model is required. In this paper, we propose a new method for estimating a default rate. It is a Bayesian approach using the power prior which allows for incorporating of historical data to estimate the default rate. Inference on current data could be more reliable if there exist similar data based on previous studies. Ibrahim and Chen (2000) utilize these data to characterize the power prior. It allows for incorporating of historical data to estimate the parameters in the models. We demonstrate our methodologies with a real data set regarding SOHO data and also perform a simulation study.

Inference on current data could be more reliable if there exist similar data based on previous studies. Ibrahim and Chen (2000) utilize these data to characterize the power prior. The power prior is constructed by raising the likelihood function of the historical data to the power $a_o$, where $0\;{\le}\;a_o\;{\le}\;1$. The power prior is a useful informative prior in Bayesian inference. However, for model selection or model comparison problems, the propriety of the power prior is one of the critical issues. In this paper, we suggest two joint power priors for the power law process and show that they are proper under some conditions. We demonstrate our results with a real dataset and some simulated datasets.

We consider the problem of estimating the system reliability noninformative priors when both stress and strength follow generalized gamma distributions. We first derive Jeffreys' prior, group ordering reference priors, and matching priors. We investigate the propriety of posterior distributions and provide marginal posterior distributions under those noninformative priors. We also examine whether the reference priors satisfy the probability matching criterion.

A momentum threshold autoregressive (MTAR) model, a nonlinear autoregressive model, is analyzed in a Bayesian framework. Parameter estimation in the presence of missing data is done by using Markov chain Monte Carlo methods. We also propose simple Bayesian test procedures for asymmetry and unit roots. The proposed method is applied to a set of Korea unemployment rate data and reveals evidence for asymmetry and a unit root.