/pdf/bayesian-survival-analysis-312ols7wam.pdf

Bayesian Survival Analysis.

박응격 저, 「行政學講義」(서울 : 박영사, 1993)

The Full Bayesian Signicance Test, FBST, is extensively reviewed. Its test statistic, a genuine Bayesian measure of evidence, is discussed in detail. Its behavior in some problems of statistical inference like testing for independence in contingency tables is discussed.

/pdf/can-a-significance-test-be-genuinely-bayesian-13oc0xgydt.pdf

Can a Significance Test Be Genuinely Bayesian

Mining big data using parsimonious factor, machine learning, variable selection and shrinkage methods

A number of recent studies in the economics literature have focused on the usefulness of factor models in the context of prediction using "big data". We add to this literature by analyzing whether "big data" are useful for modelling low frequency macroeconomic variables such as unemployment, inflation and GDP. In particular, we analyze the predictive benefits associated with the use of principal component analysis (PCA), independent component analysis (ICA), and sparse principal component analysis (SPCA). We also evaluate machine learning, variable selection and shrinkage methods, including bagging, boosting, ridge regression, least angle regression, the elastic net, and the non-negative garotte. Our approach is to carry out a forecasting "horse-race" using prediction models constructed using a variety of model specification approaches, factor estimation methods, and data windowing methods, in the context of the prediction of 11 macroeconomic variables relevant for monetary policy assessment. In many instances, we find that various of our benchmark models, including autoregressive (AR) models, AR models with exogenous variables, and (Bayesian) model averaging, do not dominate specifications based on factor-type dimension reduction combined with various machine learning, variable selection, and shrinkage methods (called "combination" models). We find that forecast combination methods are mean square forecast error (MSFE) "best" for only 3 of 11 variables when the forecast horizon, h=1, and for 4 variables when h=3 or 12. Additionally, non-PCA type factor estimation methods yield MSFE-best predictions for 9 of 11 variables when h=1, although PCA dominates at longer horizons. Interestingly, we also find evidence of the usefulness of combination models for approximately 1/2 of our variables, when h>1. Most importantly, we present strong new evidence of the usefulness of factor based dimension reduction, when utilizing "big data" for macroeconometric forecasting.

https://papers.ssrn.com/sol3/Delivery.cfm/SSRN_ID2747473_code76028.pdf?abstractid=2747473&mirid=4

Mining Big Data Using Parsimonious Factor, Machine Learning, Variable Selection and Shrinkage Methods

Sparse latent multi-factor models have been used in many exploratory and predictive problems with high-dimensional multivariate observations. Because of concerns with identifiability, the latent factors are almost always assumed to be linearly related to measured feature variables. Here we explore the analysis of multi-factor models with different structures of interactions between latent factors, including multiplicative effects as well as a more general framework for nonlinear interactions introduced via the Gaussian Process. We utilize sparsity priors to test whether the factors and interaction terms have significant effect. The performance of the models is evaluated through simulated and real data applications in genomics. Variation in the number of copies of regions of the genome is a well-known and important feature of most cancers. We examine interactions between factors directly associated with different chromosomal regions detected with copy number alteration in breast cancer data. In this context, significant interaction effects for specific genes suggest synergies between duplications and deletions in different regions of the chromosome.

https://projecteuclid.org/download/pdfview_1/euclid.aoas/1372338468

Sparse latent factor models with interactions: Analysis of gene expression data

In this study we introduce a likelihood-based method, via the Weibull and piecewise exponential distributions, capable of accommodating the dependence between failure and censoring times. The methodology is developed for the analysis of clustered survival data and it assumes that failure and censoring times are mutually independent conditional on a latent frailty. The dependent censoring mechanism is accounted through the frailty effect and this is accomplished by means of a key parameter accommodating the correlation between failure and censored observations. The full specification of the likelihood in our work simplifies the inference procedures with respect to Huang and Wolfe since it reduces the computation burden of working with the profile likelihood. In addition, the assumptions made for the baseline distributions lead to models with continuous survival functions. In order to carry out inferences, we devise a Monte Carlo EM algorithm. The performance of the proposed models is investigated through a simulation study. Finally, we explore a real application involving patients from the Dialysis Outcomes and Practice Patterns Study observed between 1996 and 2015.

An approach to model clustered survival data with dependent censoring.

https://repositorio.ufmg.br/bitstream/1843/RAOA-BB9KHB/1/disserta__o_gil__guilherme.pdf

Spatial statistical methods applied to the 2015 Brazilian energy distribution benchmarking model: Accounting for unobserved determinants of inefficiencies

In this paper we analyze the fraction of non-disjunction in Meiosis I assuming reference (non-informative) priors. We consider Jeffreys's approach to built a non-informative prior (Jeffreys's prior) for the fraction of non-disjunction in Meiosis I. We prove that Jeffreys's prior is a proper distribution. We perform Monte Carlo studies in order to compare Bayes estimates obtained assuming Jeffreys's and uniform priors. We consider full Bayesian significance test (FBST) and Bayes factor (BF) for testing precise hypothesis on the fraction of non-disjunction in Meiosis I. The ultimate goal of this paper is to compare these two test procedures through simulation studies using both prior specifications. An application to Down Syndrome data is also presented.

/pdf/testing-and-estimating-the-non-disjunction-fraction-in-4ubiu421sk.pdf

Testing and estimating the non-disjunction fraction in Meiosis I using reference priors.

A common problem in the analysis of gene expression microarray data is the identification of groups of features that are coherently expressed. For example, one often wishes to know whether a group of genes, clustered because of correlation in one data set, are still highly co-expressed in another data set. Alternatively, for some expression array platforms there are many, relatively short probes for each gene of interest. In this case, it is possible that a given probe is not measuring its targeted gene, but rather a different gene with a similar region (called cross-hybridization). Accurate detection of the collection of probe sets (groups of probes targeting the same gene) which demonstrate highly coherent expression patterns is the best approach to the identification of which genes are present in the sample. We develop a Bayesian Factor Model (BFM) to address the general problem of detection of coherent patterns in gene expression data sets. We compare our method to “state of the art” methods for the identification of expressed genes in both synthetic and real data sets, and the results indicate that the BFM outperforms the other procedures for detecting transcripts. We also demonstrate the use of factor analysis to identify the presence/absence status of gene modules (groups of coherently expressed genes). Variation in the number of copies of regions of the genome is a well known and important feature of most cancers. We examine a group of genes, representative of Copy Number Alteration (CNA) in breast cancer, then identify the presence/absence of CNA in this region of the genome for other cancers. Coherent patterns can also be evaluated in high-throughput sequencing data, a novel technology to measure gene expression. We analyze this type of data via factor model and examine the detection calls in terms of read mapping uncertainty.

Vinícius Diniz Mayrink

Papers

Sparse latent factor models with interactions: Analysis of gene expression data

An approach to model clustered survival data with dependent censoring.

Spatial statistical methods applied to the 2015 Brazilian energy distribution benchmarking model: Accounting for unobserved determinants of inefficiencies

Testing and estimating the non-disjunction fraction in Meiosis I using reference priors.

Bayesian factor models for the detection of coherent patterns in gene expression data