Diffusive nested sampling

Knowledge of the Earth's interior relies on indirect information collected at or near the surface. Typically, data do not uniquely constrain the subsurface properties and are contaminated by noise, and therefore the solution of ill-posed inverse problems is required. Geophysicists have traditionally addressed such problems through deterministic approaches, seeking a single 'best-fitting' model defined by optimality criteria that reflect our understanding of the problem at hand. In recent decades, Bayesian approaches have become increasingly common, as they characterise the posterior probability distribution of the model conditioned on the observations, thereby quantifying uncertainty.In this context, Markov chain Monte Carlo (MCMC) methods have emerged as a crucial tool as they allow sampling from posterior distributions of arbitrary complexity. At the core of many MCMC algorithms lies the Metropolis-Hastings scheme. This combines a proposal distribution with a probabilistic acceptance criterion to construct a Markov chain that has the desired target distribution as its stationary distribution. The algorithm is versatile as it rests on mild technical conditions on the proposal, and is thus widely adopted across a broad range of geoscientific inference problems. Yet when the parameter space is large or the forward models are computationally expensive---both common scenarios in geophysical applications---it can become inefficient, resulting in poor chain mixing and slow convergence to the target (posterior) distribution. These challenges underscore the importance of effective proposal mechanisms.In this presentation, we introduce a novel approach to designing Metropolis-Hastings proposals based on adaptive transport maps. The framework is inspired by recent developments from the field of Applied Mathematics linking Bayesian inference with Optimal Transport theory. The idea is to find a monotone, nonlinear transformation to recast a (complex) target probability distribution into a (simpler) reference distribution that is more amenable to standard MCMC steps. Our key contribution is to parameterise these transformations using invertible neural networks, ensuring monotonicity while gaining the flexibility and expressiveness that neural architectures afford.The proposed method progresses iteratively. We begin with a standard sampling strategy (e.g., a random-walk Metropolis) to obtain initial draws from the target distribution. These samples inform the training of an invertible neural network that learns to map from the target to a simpler reference distribution, specifically a standard (multivariate) Gaussian. Subsequent proposals are then generated in the reference space, either as global independence moves or local perturbations, and are accepted or rejected following a suitably modified Metropolis-Hastings criterion. As more samples accumulate, the network's parameters are updated, improving overall sampling efficiency. Ultimately, the approach yields not only an ensemble of samples representative of the desired target distribution---just as in standard MCMC---but also a compact, learned representation of it in the network's weights.We illustrate the proposed paradigm both theoretically and through examples. Preliminary results indicate that transport-map-enhanced MCMC has the potential to significantly accelerate Bayesian sampling across a range of applications.

Density-estimation likelihood-free inference (DELFI) has recently been\nproposed as an efficient method for simulation-based cosmological parameter\ninference. Compared to the standard likelihood-based Markov Chain Monte Carlo\n(MCMC) approach, DELFI has several advantages: it is highly parallelizable,\nthere is no need to assume a possibly incorrect functional form for the\nlikelihood and complicated effects (e.g the mask and detector systematics) are\neasier to handle with forward models. In light of this, we present two DELFI\npipelines to perform weak lensing parameter inference with lognormal\nrealizations of the tomographic shear field -- using the C_l summary statistic.\nThe first pipeline accounts for the non-Gaussianities of the shear field,\nintrinsic alignments and photometric-redshift error. We validate that it is\naccurate enough for Stage III experiments and estimate that O(1000) simulations\nare needed to perform inference on Stage IV data. By comparing the second DELFI\npipeline, which makes no assumption about the functional form of the\nlikelihood, with the standard MCMC approach, which assumes a Gaussian\nlikelihood, we test the impact of the Gaussian likelihood approximation in the\nMCMC analysis. We find it has a negligible impact on Stage IV parameter\nconstraints. Our pipeline is a step towards seamlessly propagating all\ndata-processing, instrumental, theoretical and astrophysical systematics\nthrough to the final parameter constraints.\n

SummaryMarkov Chain Monte Carlo (MCMC) methods are used to sample from complicated multivariate distributions with normalizing constants that may not be computable in practice and from which direct sampling is not feasible. A fundamental problem is to determine convergence of the chains. Propp & Wilson (1996) devised a Markov chain algorithm called Coupling From The Past (CFTP) that solves this problem, as it produces exact samples from the target distribution and determines automatically how long it needs to run. Exact sampling by CFTP and other methods is currently a thriving research topic. This paper gives a review of some of these ideas, with emphasis on the CFTP algorithm. The concepts of coupling and monotone CFTP are introduced, and results on the running time of the algorithm presented. The interruptible method of Fill (1998) and the method of Murdoch & Green (1998) for exact sampling for continuous distributions are presented. Novel simulation experiments are reported for exact sampling from the Ising model in the setting of Bayesian image restoration, and the results are compared to standard MCMC. The results show that CFTP works at least as well as standard MCMC, with convergence monitored by the method of Raftery & Lewis (1992, 1996).

Time series models are often used in hydrology and meteorology studies to model streamflows series in order to make forecasting and generate synthetic series which are inputs for the analysis of complex water resources systems. In thispaper we introduce a new modeling approach for hydrologic and meteorological time series assuming a continuous distribution for the data, where both the conditional mean and conditional varianceparameters are modeled. Bayesian methods using standard MCMC (Markov Chain Monte Carlo Methods) are used to simulate samples for the joint posterior distribution of interest. Two applications to real data sets illustrate the proposedmethodology, assuming that the observations come from a normal, a gamma or a beta distribution. A first example is given by a time series of monthly averages of natural streamflows, measured in the year period ranging from1931 to 2010 in Furnas hydroelectric dam, Brazil. A second example is given with a time series of 313 air humidity data measured in a weather station of Rio Claro, a Brazilian city located in southeastern of Brazil. These applications motivate us to introduce new classes of models to analyze hydrological and meteorological time series

Background: Rapid progression in medical and health sciences have caused survival studies, where some patients have long-term survival, especially for chronic diseases such as breast cancer. Cure models can be applicable to analyze such data. Objectives: The aim of this study was to determine the risk factors associated with breast cancer, using mixture cure fraction model. Methods: We studied data for 438 patients, who were referred to cancer research center in Shahid Beheshti University of Medical Sciences. The patients were visited and treated during 1992 to 2012 and followed-up until October 2014. The data were analyzed by mixture cure fraction model based on GMW (generalized modified Weibull) distribution and inferences were obtained with Bayesian approach, using standard MCMC (Markov Chain Monte Carlo) methods. All analyses were performed, using SPSS v20 and OpenBUGS software. The significant level was considered at 0.05. Results: During the follow-up period, 75 (17.12%) deaths occurred by breast cancer and the one-year overall survival rate was 98%. Covariates such as numbers of metastatic lymph nodes and histologic grade were statistically significant. Also, the cure fraction estimation was obtained 58%. Conclusions: When some patients have a long-term survival, cure models can be an interesting model to study survival and these models estimate parameters better than the traditional models such as cox model. In this paper, the mixture cure fraction model based on GMW was fitted for analysing survival times in patients with breast cancer.

Airport air traffic is one of the most important and hardest one among all the airport data forecasts. In this paper, the Markov Chain Monte Carlo (MCMC) method of applied theory of statistics has been introduced into the aviation sector, and the discussion on airport air traffic forecast has been conducted taking Shanghai airport as the application background. MCMC method is designed to solve the problem parameters expectations where some baysian inference can not be directly calculated. This paper conducts the iterative computation using WinBUGS, EViews is used to estimate the initial iteration parameter of the regression model, and it is on this basis that the regression equation under the MCMC method is obtained.

Parameter estimation and uncertainty analysis for a watershed model

Markov chain Monte Carlo sampling using a reservoir method

The conventional Markov chain Monte Carlo (MCMC) method is limited to the selected shape and size of proposal distribution and is not easy to start when the initial proposal distribution is far away from the target distribution. To overcome these drawbacks of the conventional MCMC method, two useful improvements in MCMC method, adaptive Metropolis (AM) algorithm and delayed rejection (DR) algorithm, are attempted to be combined. The AM algorithm aims at adapting the proposal distribution by using the generated estimators, and the DR algorithm aims at enhancing the efficiency of the improved MCMC method. Based on the improved MCMC method, a Bayesian amplitude versus offset (AVO) inversion method on the basis of the exact Zoeppritz equation has been developed, with which the P- and S-wave velocities and the density can be obtained directly, and the uncertainty of AVO inversion results has been estimated as well. The study based on the logging data and the seismic data demonstrates the feasibility and robustness of the method and shows that all three parameters are well retrieved. So the exact Zoeppritz-based nonlinear inversion method by using the improved MCMC is not only suitable for reservoirs with strong-contrast interfaces and long-offset ranges but also it is more stable, accurate and anti-noise.

Markov chain Monte Carlo (MCMC) methods offer a rapid parametric approach that can test for linkage throughout the entire genome. It has an advantage similar to nonparametric methods in that the model does not have to be completely specified a priori. However, unlike nonparametric methods, there are no limitations on pedigree size and MCMC methods can also handle relatively complex pedigree structures. In addition MCMC methods can be used to carry segregation analysis in order to answer questions on the genetic components of a disease phenotype. Segregation analysis gave evidence for between two and eight alcoholism susceptibility loci, each having a modest effect on the phenotype. MCMC methods were used to map alcoholism loci using the phenotypes ALDX1 (DSM-III-R and Feighner criteria) and ALDX2 (World Health Organization diagnosis ICD-10 criteria). There was mild evidence for quantitative trait loci on chromosomes 2, 10, and 11.

Comparing the performance of different multiple imputation strategies for missing binary outcomes in cluster randomized trials: a simulation study Jinhui Ma,1–3 Parminder Raina,1,2 Joseph Beyene,1 Lehana Thabane1,3–51Department of Clinical Epidemiology and Biostatistics, McMaster University, Hamilton, ON, Canada; 2McMaster University Evidence-based Practice Center, Hamilton, ON, Canada; 3Biostatistics Unit, St Joseph's Healthcare Hamilton, Hamilton, ON, Canada; 4Centre for Evaluation of Medicines, St Joseph's Healthcare Hamilton, Hamilton, ON, Canada; 5Population Health Research Institute, Hamilton Health Sciences, Hamilton, ON, CanadaIntroduction: Although researchers have proposed various strategies to handle missing outcomes in cluster randomized trials (CRTs), limited attention has been paid to the performance of these strategies. Under the assumption of covariate-dependent missingness, the objective of this simulation study is to compare the performance of various strategies in handling missing binary outcomes in CRTs under different design settings.Methods: There are six missing data strategies investigated in this paper, which include complete case analysis, standard multiple imputation (MI) strategies using either logistic regression or Markov chain Monte Carlo (MCMC) method, within-cluster MI strategies using either logistic regression or MCMC method, and MI using logistic regression with cluster as a fixed effect. The performance of these strategies is evaluated through bias, empirical standard error, root mean squared error, and coverage probability.Results: Under the assumption of covariate-dependent missingness and applying the generalized estimating equations approach for fitting the logistic regression, it was shown that complete case analysis yields valid inferences when the percentage of missing outcomes is not large (<20%) for all the designs of CRTs considered in this paper. Standard MI strategies can be adopted when the design effect is small (variance inflation factor [VIF] ≤ 3); however, they tend to underestimate the standard error of treatment effect when the design effect is large. Within-cluster MI strategy using logistic regression is valid for imputation of missing data from CRTs, especially when the cluster size is large (>50) and the design effect is large (VIF > 3). In contrast, within-cluster MI strategy using MCMC method may yield biased estimates of treatment effect for CRTs with small cluster size (≤50). MI using logistic regression with cluster as a fixed effect may substantially overestimate the standard error of the estimated treatment effect when the intracluster correlation coefficient is small. It may also lead to biased estimated treatment effect.Conclusion: Findings from this simulation study provide researchers with quantitative evidence to guide selection of an appropriate strategy to deal with missing binary outcomes.Keywords: missing data, design effect, variance inflation factor

Clustering is a fundamental problem in statistics and machine learning, whose solution is commonly computed by the Expectation-Maximization (EM) method, which finds a locally optimal solution for an objective function called log-likelihood. Since the surface of the log-likelihood function is non convex, a stochastic search with Markov Chain Monte Carlo (MCMC) methods can help escaping locally optimal solutions. In this article, we tackle two fundamental conflicting goals: Finding higher quality solutions and achieving faster convergence. With that motivation in mind, we introduce an efficient algorithm that combines elements of the EM and MCMC methods to find clustering solutions that are qualitatively better than those found by the standard EM method. Moreover, our hybrid algorithm allows tuning model parameters and understanding the uncertainty in their estimation. The main issue with MCMC methods is that they generally require a very large number of iterations to explore the posterior of each model parameter. Convergence is accelerated by several algorithmic improvements which include sufficient statistics, simplified model parameter priors, fixing covariance matrices and iterative sampling from small blocks of the data set. A brief experimental evaluation shows promising results.

Traditional full-waveform inversion (FWI) methods only render a “best-fit” model that cannot account for uncertainties of the ill-posed inverse problem. Additionally, local optimization-based FWI methods cannot always converge to a geologically meaningful solution unless the inversion starts with an accurate background model. We seek the solution for FWI in the Bayesian inference framework to address those two issues. In Bayesian inference, the model space is directly probed by sampling methods such that we obtain a reliable uncertainty appraisal, determine optimal models, and avoid entrapment in a small local region of the model space. The solution of such a statistical inverse method is completely described by the posterior distribution, which quantifies the distributions for parameters and inversion uncertainties. To efficiently sample the posterior distribution, we introduce a sampling algorithm in which the proposal distribution is constructed by the local gradient and the diagonal approximate Hessian of the local log posterior. Our algorithm is called the gradient-based Markov chain Monte Carlo (GMCMC) method. The GMCMC FWI method can quantify inversion uncertainties with estimated posterior distribution given sufficiently long Markov chains. By directly sampling the posterior distribution, we obtain a global view of the model space. Theoretically speaking, statistical assessments do not depend on starting models. Our method is applied to the 2D Marmousi model with the frequency-domain FWI setting. Numerical results suggest that our method can be readily applied to 2D cases with affordable computational efforts.

In order to gain an understanding of the effectiveness of phylogenetic Markov chain Monte Carlo (MCMC), it is important to understand how quickly the empirical distribution of the MCMC converges to the posterior distribution. In this article, we investigate this problem on phylogenetic tree topologies with a metric that is especially well suited to the task: the subtree prune-and-regraft (SPR) metric. This metric directly corresponds to the minimum number of MCMC rearrangements required to move between trees in common phylogenetic MCMC implementations. We develop a novel graph-based approach to analyze tree posteriors and find that the SPR metric is much more informative than simpler metrics that are unrelated to MCMC moves. In doing so, we show conclusively that topological peaks do occur in Bayesian phylogenetic posteriors from real data sets as sampled with standard MCMC approaches, investigate the efficiency of Metropolis-coupled MCMC (MCMCMC) in traversing the valleys between peaks, and show that conditional clade distribution (CCD) can have systematic problems when there are multiple peaks.

As alluded to in the chapter “Linkage Analysis of Qualitative Traits”, neither the Elston–Steward algorithm nor the Lander–Green approach is amenable to genetic data from large complex pedigrees and a large number of markers. In such cases, Monte Carlo estimation methods provide a viable alternative to the exact solutions. Two types of Monte Carlo methods have been developed for linkage analysis, haplotype inference, and other kinds of genetic analysis. They are Markov chain Monte Carlo (MCMC) methods and Monte Carlo methods that are based on independent samples. Approaches based on Markov chain Monte Carlo methods are more widely applicable; there is practically no limit on the size or complexity of the pedigrees, nor on the number of markers to be considered simultaneously. In this chapter, we will review the basic principles of MCMC methods for multipoint linkage analysis with extended pedigrees. Both simulations and application to data from the Framingham study will be used to compare and contrast three MCMC software packages: LOKI, MORGAN, and SIMWALK.