Contrast trees and distribution boosting

A method for decision tree induction is presented. Given a set of predictor variablesx=(x1,x2,⋅⋅⋅,xp)and two outcome variables y and z associated with each x, the goal is to identify those values of x for which the respective distributions ofy | xandz | x, or selected properties of those distributions such as means or quantiles, are most different. Contrast trees provide a lack-of-fit measure for statistical models of such statistics, or for the complete conditional distributionpy(y | x), as a function of x. They are easily interpreted and can be used as diagnostic tools to reveal and then understand the inaccuracies of models produced by any learning method. A corresponding contrast-boosting strategy is described for remedying any uncovered errors, thereby producing potentially more accurate predictions. This leads to a distribution-boosting strategy for directly estimating the full conditional distribution of y at each x under no assumptions concerning its shape, form, or parametric representation.

Fully Bayesian Estimation of Simultaneous Regression Quantiles under Asymmetric Laplace Distribution Specification

In this paper, we are interested in estimating several quantiles simultaneously in a regression context via the Bayesian approach. Assuming that the error term has an asymmetric Laplace distribution and using the relation between two distinct quantiles of this distribution, we propose a simple fully Bayesian method that satisfies the noncrossing property of quantiles. For implementation, we use Metropolis-Hastings within Gibbs algorithm to sample unknown parameters from their full conditional distribution. The performance and the competitiveness of the underlying method with other alternatives are shown in simulated examples.

Conditional expectation estimation through attributable components

A general methodology is proposed for the explanation of variability in a quantity of interest x in terms of covariates z = (z1, …, zL). It provides the conditional mean $\bar{x}(z)$ as a sum of components, where each component is represented as a product of non-parametric one-dimensional functions of each covariate zl that are computed through an alternating projection procedure. Both x and the zl can be real or categorical variables; in addition, some or all values of each zl can be unknown, providing a general framework for multi-clustering, classification and covariate imputation in the presence of confounding factors. The procedure can be considered as a preconditioning step for the more general determination of the full conditional distribution $\boldsymbol{\rho}(x|z) $ through a data-driven optimal-transport barycenter problem. In particular, just iterating the procedure once yields the second order structure (i.e. the covariance) of $\boldsymbol{\rho}(x|z) $. The methodology is illustrated through examples that include the explanation of variability of ground temperature across the continental United States and the prediction of book preference among potential readers.

Mixture Normal Distribution for Gibbs Sampler and its Application in the Surface of Single Crystal

Gibbs sampler is widely used in Bayesian analysis. But it is often difficult to sample from the full conditional distribution, and this hardly weakens the efficiency of Gibbs sampler. In this paper, we propose to use mixture normal distribution for Gibbs sampler. The mixture normal distribution can approximate the target distribution. So carrying more information from target distribution, the mixture normal distribution tremendously improves the efficiency of Gibbs sampler. Further more, combining with mixture normal method, Hit-and-Run algorithm can also get more efficient sampling results. Simulation results show that Gibbs sampler with mixture normal distribution outperforms other sampling algorithms. The Gibbs sampler with mixture normal distribution can also be applied to explorer the surface of single crystal.