Generalized Markov-Polya urn models with predetermined strategies

By using combinatorial methods involving lattice path combinatorics, three generalized probability models dependent on predetermined strategies have been obtained with the help of urn models.The models have been developed with the help of a sampling scheme which unifies both, the binomial and the inverse binomial sampling schemes. These models generate a number of important discrete probability distributions both as particular cases and as limiting cases. Recurrence relations among the moments of the models have also been obtained.     

The two-color urn model: recursions for the moments

For the two-color reinforcement-depletion urn model, with balancing reinforcement and depletion held constant over cycles, a recursive formula is given from which all factorial moments (for white balls, for example) can be determined. When the reinforcement of each color is positive, the stationary distribution of white balls (infinite number of cycles) turns out to be determined by three parameters. namely (i) the total number of balls in the urn, (ii) the richness of the reinforcement, or ratio of white ball reinforcement to total reinforcement, and (iii) the size of the white ball reinforcement. In addition, the distribution mimics the binomial (with less variance and skewness (√β₁:) ) and from formulas for the exact first four moments rapidly approaches normality. On the basis of the few cases studied, an approximating Gram-Charlier distribution with a binomial nucleus is only moderately successful     

Interim analysis of clinical trials based on urn models

Clinical trials usually involve efficient and ethical objectives such as maximizing the power and minimizing the total failure number. Interim analysis is now a standard technique in practice to achieve these objectives. Randomized urn models have been extensively studied in the literature. In this paper, we propose to perform interim analysis on clinical trials based on urn models and study its properties. We show that the urn composition, allocation of patients and parameter estimators can be approximated by a joint Gaussian process. Consequently, sequential test statistics of the proposed procedure converge to a Brownian motion in distribution and the sequential test statistics asymptotically satisfy the canonical joint distribution defined in Jennison & Turnbull (Jennison & Turnbull 2000. Group Sequential Methods with Applications to Clinical Trials, Chapman and Hall/CRC). These results provide a solid foundation and open a door to perform the interim analysis on randomized clinical trials with urn models in practice. Furthermore, we demonstrate our proposal through examples and simulations by applying sequential monitoring and stochastic curtailment techniques. The Canadian Journal of Statistics 40: 550–568; 2012 © 2012 Statistical Society of Canada     

A goodness-of-fit test using randomly selected cell boundaries

This paper describes a technique for testing for the goodness-of-fit of a set of observations on a random variable to a hypothesized distribution. The cell boundaries needed for partitioning the range of the random variable are randomly chosen using the computer, thus eliminating an arbitrary choice by the analyst. The power of this test procedure is compared to the test procedure using equal expected cell frequencies     

Computing the probability of hash table/urn overflow

We analyze the probability of a random distribution of n balls into m urns of size b resulting in no overflows. This solves the computational problem associated with a classical combinatorial extreme-value distribution. The problem arose during the analysis of a technique, called perfect hashing, for organizing data in computer files. The results and techniques presented can be used to solve several problems in the analysis of hashing techniques     

Normal approximation in an urn model with indistinguishable balls

For the Bose-Einstein Statistics, where n indistinguishable balls are distributed in m urns such that all the arrangements are equally likely, define the random variables

M_k = number of urns containing exactly k balls each;

N_k = number of urns containing at least k balls each.

We consider the approximation of the distributions of M_k and N_k by suitable normal distributions, for large but finite m. Estimates are found for the error in the approximation to both the probability mass function and the distribution function in each case. These results apply also to the alternative model where no urn is allowed to be empty. The results are illustrated by some numerical examples.     

Extending the family of Bayesian bootstraps and exchangeable urn schemes

A Bayesian bootstrap for a finite population with censored observations is introduced. It is shown to reduce to the finite population Bayesian bootstrap if there is no censoring and to reduce to the censored data Bayesian bootstrap for a large population. A class of general urn schemes for simulating exchangeable sequences of variables is introduced which is connected to the bootstrap method.     

Some extensions and applications of Kriz's (1972) urn model

A summary of Kriz's (1972) paper concerning an urn model for the PMP distribution is given. Extensions of the model are made, considering cases in which the urns are not drawn from independently. Possible applications of the extended model in some epidemiological situations are suggested, and a simple computer algorithm for simulating the distribution is given. Results of a simple simulation based on May's (1986) paper are given.     

An urn model to construct an efficient test procedure for response adaptive designs

We study the statistical performance of different tests for comparing the mean effect of two treatments. Given a reference classical test \({\mathcal {T}}_0\), we determine which sample size and proportion allocation guarantee to a test \({\mathcal {T}}\), based on response-adaptive design, to be better than \({\mathcal {T}}_0\), in terms of (a) higher power and (b) fewer subjects assigned to the inferior treatment. The adoption of a response-adaptive design to implement the random allocation procedure is necessary to ensure that both (a) and (b) are satisfied. In particular, we propose to use a Modified Randomly Reinforced Urn design and we show how to perform the model parameters selection for the purpose of this paper. Then, the opportunity of relaxing some assumptions on treatment response distributions is presented. Results of simulation studies on the test performance are reported and a real case study is analyzed.     

Modified spline regression based on randomly right-censored data: A comparative study

ABSTRACT

In this paper, we propose modified spline estimators for nonparametric regression models with right-censored data, especially when the censored response observations are converted to synthetic data. Efficient implementation of these estimators depends on the set of knot points and an appropriate smoothing parameter. We use three algorithms, the default selection method (DSM), myopic algorithm (MA), and full search algorithm (FSA), to select the optimum set of knots in a penalized spline method based on a smoothing parameter, which is chosen based on different criteria, including the improved version of the Akaike information criterion (AICc), generalized cross validation (GCV), restricted maximum likelihood (REML), and Bayesian information criterion (BIC). We also consider the smoothing spline (SS), which uses all the data points as knots. The main goal of this study is to compare the performance of the algorithm and criteria combinations in the suggested penalized spline fits under censored data. A Monte Carlo simulation study is performed and a real data example is presented to illustrate the ideas in the paper. The results confirm that the FSA slightly outperforms the other methods, especially for high censoring levels.     

A compound Rayleigh survival model and its application to randomly censored data

Received: May 5, 1999; revised version: June 15, 2000     

A class of nonparametric bivariate survival function estimators for randomly censored and truncated data

This paper proposes a class of nonparametric estimators for the bivariate survival function estimation under both random truncation and random censoring. In practice, the pair of random variables under consideration may have certain parametric relationship. The proposed class of nonparametric estimators uses such parametric information via a data transformation approach and thus provides more accurate estimates than existing methods without using such information. The large sample properties of the new class of estimators and a general guidance of how to find a good data transformation are given. The proposed method is also justified via a simulation study and an application on an economic data set.     

A q-Pólya urn model and the q-Pólya and inverse q-Pólya distributions

A q-Pólya urn model is introduced by assuming that the probability of drawing a white ball at a drawing varies geometrically, with rate q, both with the number of drawings and the number of white balls drawn in the previous drawings. Then, the probability mass functions and moments of (a) the number of white balls drawn in a specific number of drawings and (b) the number of black balls drawn until a specific number of white balls are drawn are derived. These two distributions turned out to be q-analogs of the Pólya and the inverse Pólya distributions, respectively. Also, the limiting distributions of the q-Pólya and the inverse q-Pólya distributions, as the number of balls in the urn tends to infinity, are shown to be a q-binomial and a negative q-binomial distribution, respectively. In addition, the positive or negative q-hypergeometric distribution is obtained as conditional distribution of a positive or negative q-binomial distribution, given its sum with another positive or negative q-binomial distribution, independent of it.     

A polar coordinate transformation for estimating bivariate survival functions with randomly censored and truncated data

This paper proposes a new estimator for bivariate distribution functions under random truncation and random censoring. The new method is based on a polar coordinate transformation, which enables us to transform a bivariate survival function to a univariate survival function. A consistent estimator for the transformed univariate function is proposed. Then the univariate estimator is transformed back to a bivariate estimator. The estimator converges weakly to a zero-mean Gaussian process with an easily estimated covariance function. Consistent truncation probability estimate is also provided. Numerical studies show that the distribution estimator and truncation probability estimator perform remarkably well.     

Information attainable in some randomly incomplete data models

The Fisher information is intricately linked to the asymptotic (first-order) optimality of maximum likelihood estimators for parametric complete-data models. When data are missing completely at random in a multivariate setup, it is shown that information in a single observation is well-defined and it plays the same role as in the complete-data model in characterizing the first-order asymptotic optimality properties of associated maximum likelihood estimators; computational aspects are also thoroughly appraised. As an illustration, the logistic regression model with incomplete binary responses and an incomplete categorical covariate is worked out.     

Bayesian inference for the randomly censored Weibull distribution

In this paper, we consider the Bayesian inference of the unknown parameters of the randomly censored Weibull distribution. A joint conjugate prior on the model parameters does not exist; we assume that the parameters have independent gamma priors. Since closed-form expressions for the Bayes estimators cannot be obtained, we use Lindley's approximation, importance sampling and Gibbs sampling techniques to obtain the approximate Bayes estimates and the corresponding credible intervals. A simulation study is performed to observe the behaviour of the proposed estimators. A real data analysis is presented for illustrative purposes.     

Estimation in Maxwell distribution with randomly censored data

In many practical situations, complete data are not available in lifetime studies. Many of the available observations are right censored giving survival information up to a noted time and not the exact failure times. This constitutes randomly censored data. In this paper, we consider Maxwell distribution as a survival time model. The censoring time is also assumed to follow a Maxwell distribution with a different parameter. Maximum likelihood estimators and confidence intervals for the parameters are derived with randomly censored data. Bayes estimators are also developed with inverted gamma priors and generalized entropy loss function. A Monte Carlo simulation study is performed to compare the developed estimation procedures. A real data example is given at the end of the study.     

Estimation in two sample randomly truncated scale model

This paper studies a class of Cramér–von Mises type minimum distance estimators of the scale parameter in the two sample randomly left truncated scale models. The proposed class of estimators includes an analogue of the well-known Hodges–Lehmann estimator. The paper proves the asymptotic normality of these estimators under mild conditions. It also contains a real data application and a simulation study making a comparison of some of the estimators in the class with the ratio of the two means.