Empiric bayes estimation of binomial probability

An empirical Bayes estimator of a binomial parameter, based on orthogonal polynomials on (0,1), is introduced. The resulting estimator of the prior density is asymptotically optimal. The method allows one to combine Bayes and empiric Bayes methods with smoothing in a natural way.     

Optimal Viterbi Bayesian predictive classification for data from finite alphabets

A family of Viterbi Bayesian predictive classifiers has been recently popularized for speech recognition applications with continuous acoustic signals modeled by finite mixture densities embedded in a hidden Markov framework. Here we generalize such classifiers to sequentially observed data from multiple finite alphabets and derive the optimal predictive classifier under exchangeability of the emitted symbols. We demonstrate that the optimal predictive classifier which learns from unlabelled test items improves considerably upon marginal maximum a posteriori rule in the presence of sparse training data. It is shown that the learning process saturates when the amount of test data tends to infinity, such that no further gain in classification accuracy is possible upon arrival of new test items in the long run.     

Have I seen you before? Principles of Bayesian predictive classification revisited

A general inductive Bayesian classification framework is considered using a simultaneous predictive distribution for test items. We introduce a principle of generative supervised and semi-supervised classification based on marginalizing the joint posterior distribution of labels for all test items. The simultaneous and marginalized classifiers arise under different loss functions, while both acknowledge jointly all uncertainty about the labels of test items and the generating probability measures of the classes. We illustrate for data from multiple finite alphabets that such classifiers achieve higher correct classification rates than a standard marginal predictive classifier which labels all test items independently, when training data are sparse. In the supervised case for multiple finite alphabets the simultaneous and the marginal classifiers are proven to become equal under generalized exchangeability when the amount of training data increases. Hence, the marginal classifier can be interpreted as an asymptotic approximation to the simultaneous classifier for finite sets of training data. It is also shown that such convergence is not guaranteed in the semi-supervised setting, where the marginal classifier does not provide a consistent approximation.     

Infant mortality by cause: A comparison of underlying and multiple cause designations

Through the use of unique data for the 1980–1982 birth cohorts in the State of Florida, USA, the present analysis examines the extent of variability in the social, economic, and demographic characteristics of decedents when grouped by detailed cause of infant death, across four models for identifying cause of death. The analysis first compares cause-of-death-specific infant, neonatal, and postneonatal mortality between each of the four cause-of-death models. Next, interest shifts to an examination of the variability among decedents, specific to cause of death, in a range of background, proximate, and immediate determinants of infant health and survival. Variability is evident in cause-specific mortality rates as well as in decedent characteristics across the cause-of-death models. These findings suggest that more attention be given to the mode of identifying cause of death in studies of infant mortality.     

Stouffer’s Model Of Migration: A Comparison Of Interstate And Metropolitan Flows

Demography - The objectives of this paper are (1) to evaluate the accuracy of the intervening opportunities-competing migrants hypothesis (Stouffer, 1960) in estimating 1955-1960 interstate...     

Individual outpatient treatment of pathological gambling

Recovery from pathological gambling is viewed as a process whereby the pathological gambler chooses to lose an addiction to gambling and maintains that choice while mourning the loss of the gambling. The self-induced and self-escalating addictive crisis leading up to that choice is explored, as is the crisis caused by the loss of the gambling itself. It is proposed that gambling provides the gambler with action, a method of dealing with stress aand avoiding unpleasant affect, and a variety of social, psychological, and existential benefits. The loss of gambling is seen as a complicated and significant one which elicits grief responses similar to those seen in response to other types of major loss. It is proposed that in the individual outpatient treatment of pathological gamblers, the therapist helps the recovering gambler to accept the loss of the gambling and learn to live a rewarding life without it. Four phases of treatment are suggested which correspond to stages of grieving and accepting the loss of the gambling.