• Journal for History of Mathematics
• /
• v.11 no.1
• /
• pp.68-77
• /
• 1998

There are two approach methods widely in statistical inferences. First is sampling theory methods and the other is Bayesian methods. In this paper, we will introduce the most basic differences of the two approach methods. Especially, we investigate and introduce the historical origin of Bayesian methods in Statistical inferences which is currently used. Also, we introduce the some characteristics of sampling theory method and Bayesian methods.

• Journal of KIISE:Databases
• /
• v.37 no.6
• /
• pp.285-291
• /
• 2010

In this paper, we propose a new paradigm of weighting methods for naive Bayesian learning. We propose more fine-grained weighting methods, called value weighting method, in the context of naive Bayesian learning. While the current weighting methods assign a weight to an attribute, we assign a weight to an attribute value. We develop new methods, using Kullback-Leibler function, for both value weighting and feature weighting in the context of naive Bayesian. The performance of the proposed methods has been compared with the attribute weighting method and general naive bayesian. The proposed method shows better performance in most of the cases.

• Proceedings of the Korean Statistical Society Conference
• /
• 2005.11a
• /
• pp.59-65
• /
• 2005

We propose five Bayesian methods to estimate the cell expectation in an incomplete multi-way categorical table with nonignorable nonresponse mechanism. We study 3 Bayesian methods which were previously applied to one-way categorical tables. We extend them to multi-way tables and, in addition, develop 2 new Bayesian methods for multi-way categorical tables. These five methods are distinguished by different priors on the cell probabilities: two of them have the priors determined only by information of respondents; one has a constant prior; and the remaining two have priors reflecting the difference in the response mechanisms between respondent and non-respondent. We also compare the five Bayesian methods using a categorical data for a prospective study of pregnant women.

• Journal of Environmental Science International
• /
• v.28 no.1
• /
• pp.19-35
• /
• 2019

This study investigates the performance of four Bayesian methods, Random Walk Metropolis (RWM), Hit-And-Run Metropolis (HARM), Adaptive Mixture Metropolis (AMM), and Population Monte Carlo (PMC), for estimating the parameters and uncertainties of probability rainfall distribution, and the results are compared with those of conventional parameter estimation methods; namely, the Method Of Moment (MOM), Maximum Likelihood Method (MLM), and Probability Weighted Method (PWM). As a result, Bayesian methods yield similar or slightly better results in parameter estimations compared with conventional methods. In particular, PMC can reduce parameter uncertainty greatly compared with RWM, HARM, and AMM methods although the Bayesian methods produce similar results in parameter estimations. Overall, the Bayesian methods produce better accuracy for scale parameters compared with the conventional methods and this characteristic improves the accuracy of probability rainfall. Therefore, Bayesian methods can be effective tools for estimating the parameters and uncertainties of probability rainfall distribution in hydrological practices, flood risk assessment, and decision-making support.

• Communications for Statistical Applications and Methods
• /
• v.24 no.6
• /
• pp.543-559
• /
• 2017

Tree-based regression and classification ensembles form a standard part of the data-science toolkit. Many commonly used methods take an algorithmic view, proposing greedy methods for constructing decision trees; examples include the classification and regression trees algorithm, boosted decision trees, and random forests. Recent history has seen a surge of interest in Bayesian techniques for constructing decision tree ensembles, with these methods frequently outperforming their algorithmic counterparts. The goal of this article is to survey the landscape surrounding Bayesian decision tree methods, and to discuss recent modeling and computational developments. We provide connections between Bayesian tree-based methods and existing machine learning techniques, and outline several recent theoretical developments establishing frequentist consistency and rates of convergence for the posterior distribution. The methodology we present is applicable for a wide variety of statistical tasks including regression, classification, modeling of count data, and many others. We illustrate the methodology on both simulated and real datasets.

• Communications for Statistical Applications and Methods
• /
• v.23 no.6
• /
• pp.445-466
• /
• 2016

Nonparametric Bayesian methods have seen rapid and sustained growth over the past 25 years. We present a gentle introduction to the methods, motivating the methods through the twin perspectives of consistency and false consistency. We then step through the various constructions of the Dirichlet process, outline a number of the basic properties of this process and move on to the mixture of Dirichlet processes model, including a quick discussion of the computational methods used to fit the model. We touch on the main philosophies for nonparametric Bayesian data analysis and then reanalyze a famous data set. The reanalysis illustrates the concept of admissibility through a novel perturbation of the problem and data, showing the benefit of shrinkage estimation and the much greater benefit of nonparametric Bayesian modelling. We conclude with a too-brief survey of fancier nonparametric Bayesian methods.

• Communications for Statistical Applications and Methods
• /
• v.4 no.1
• /
• pp.185-191
• /
• 1997

Bayesian estimation of multinomial and Poisson parameters under starshped restriction is considered. Most Bayesian estimations in order restricted statistical inference require the high-dimensional integration which is very difficult to evaluate. Monte Carlo integration and Gibbs sampling are among alternative methods. The Bayesian estimation considered in this paper requires only evaluation of incomplete beta functions which are extensively tabulated.

• Communications for Statistical Applications and Methods
• /
• v.24 no.6
• /
• pp.561-581
• /
• 2017

Bayesian statistics can play a key role in the design and analysis of clinical trials and this has been demonstrated for medical device trials. By 1995 Bayesian statistics had been well developed and the revolution in computing powers and Markov chain Monte Carlo development made calculation of posterior distributions within computational reach. The Food and Drug Administration (FDA) initiative of Bayesian statistics in medical device clinical trials, which began almost 20 years ago, is reviewed in detail along with some of the key decisions that were made along the way. Both Bayesian hierarchical modeling using data from previous studies and Bayesian adaptive designs, usually with a non-informative prior, are discussed. The leveraging of prior study data has been accomplished through Bayesian hierarchical modeling. An enormous advantage of Bayesian adaptive designs is achieved when it is accompanied by modeling of the primary endpoint to produce the predictive posterior distribution. Simulations are crucial to providing the operating characteristics of the Bayesian design, especially for a complex adaptive design. The 2010 FDA Bayesian guidance for medical device trials addressed both approaches as well as exchangeability, Type I error, and sample size. Treatment response adaptive randomization using the famous extracorporeal membrane oxygenation example is discussed. An interesting real example of a Bayesian analysis using a failed trial with an interesting subgroup as prior information is presented. The implications of the likelihood principle are considered. A recent exciting area using Bayesian hierarchical modeling has been the pediatric extrapolation using adult data in clinical trials. Historical control information from previous trials is an underused area that lends itself easily to Bayesian methods. The future including recent trends, decision theoretic trials, Bayesian benefit-risk, virtual patients, and the appalling lack of penetration of Bayesian clinical trials in the medical literature are discussed.

• Communications for Statistical Applications and Methods
• /
• v.10 no.3
• /
• pp.731-740
• /
• 2003

Schervish(1996) and Lavine and Schervish(1999) have shown that the classical P-values and the Bayes factors fail to achieve the so-called coherence property, respectively. In this paper, we propose a new type of Bayesian P-value, namely the type LR Bayesian P-value, satisfying the coherence property. The proposed Bayesian P-values are very easy to use with since they are simple functions of likelihood ratio. Their performances are discussed and compared with those of other methods under several situations.

• Communications for Statistical Applications and Methods
• /
• v.6 no.1
• /
• pp.181-191
• /
• 1999

We consider Bayesian models allow multiple grouping of parameters for the normal means estimation problem. In particular, we consider a typical Bayesian hierarchical approach based on thepartial exchangeability where the components within a subgroup are exchangeable, but the different subgroups are not. We discuss implementation of such Bayesian procedures via Gibbs sampling. We illustrate the proposed methods with numerical examples.