Tuesday, September 20, 3.15 pm, 2022. Seminar in Statistics.

Metric cluster analysis of molecular biology data for systems identification of gene regulatory networks
James E. Blevins, Department of Statistics, Uppsala University
Abstract: download
Location: Alan Turing.

Tuesday, October 11, 3.15 pm, 2022. Seminar in Mathematical Statistics.

Nonparametric Finite Mixture: Applications in Contaminated Trials
Solomon W. Harrar, University of Kentucky, USA
Abstract:Investigating the differential effect of treatments in groups defined by patient characteristics is of paramount importance in personalized medicine. Group membership is typically determined by diagnostic devices or biomarkers, but such tools are not perfectly accurate. The impact of diagnostic misclassification or contamination in statistical inference has received only little attention in the literature. This work addresses the problem in a fully nonparametric setting. Nonparametric finite mixture is proposed for estimating and testing of meaningful yet nonparametric treatment effects. Consistent estimators and asymptotic distributions are provided for the misclassification error rates as well as treatment effects. Numerical examples show significant advantages of the proposed method in terms of bias reduction, coverage probability and power. The application of the proposed method is illustrated with data from asthma and sleep deprivation studies.
Location: Hopningspunkten.

Tuesday, October 25, 3.15 pm, 2022. Seminar in Statistics.

Data driven orthogonal basis selection for functional data analysis
Krzysztof Podgórski, Department of Statistics, Lund University
Abstract: Functional data analysis is typically performed in two steps: first, functionally representing discrete observations, and then applying functional methods, such as the functional principal component analysis, to the so-represented data. While the initial choice of a functional representation may have a significant impact on the second phase of the analysis, this issue has not gained much attention in the past. Typically, a rather ad hoc choice of some standard basis such as Fourier, wavelets, splines, etc. is used for the data transforming purpose. To address this important problem, we present its mathematical formulation, demonstrate its importance, and propose a data-driven method of functionally representing observations. The method chooses an initial functional basis by an efficient placement of the knots. A simple machine learning style algorithm is utilized for the knot selection and recently introduced orthogonal spline bases - splinets - are eventually taken to represent the data. The benefits are illustrated by examples of analyses of sparse functional data. Work is joint with Rani Basna, Hiba Nassar.
Location: Alan Turing.

Tuesday, November 15, 3.15 pm, 2022. Seminar in Statistics.

Brownian motion, bridges and Bayesian inference in BHV tree space
Will Woodman, School of Mathematics, Statistics and Physics, Newcastle University
Abstract: Analysis of genome data often produces data sets of phylogenetic trees. We introduce distributions on the space of phylogenetic trees (called tree space) parameterized by a location parameter and dispersion parameter obtained by executing Brownian motion from a fixed starting point in tree space (determined by the location parameter) for a certain time duration (corresponding to the dispersion parameter). Approximate samples can be drawn from such distributions by simulating random walks. Fitting these distributions to samples of points in tree space is challenging, as the likelihood cannot be evaluated directly. We first describe a Bayesian approach to parameter inference that uses an approximate likelihood calculated by forward simulating random walks. We illustrate the inference on data sets simulated under the multispecies coalescent model. We then describe an approach to parameter inference, which side-steps the difficulty of evaluating the likelihood by directly sampling random walk trajectories between the location parameter point and the data points, and then marginalising over these trajectories. This relies on an efficient algorithm for proposing random walk trajectories between fixed start and end points in tree space, which are often referred to as bridges. The method is illustrated on simulated data sets.
Location: John von Neumann (CHANGE OF ROOM!) .

Tuesday, November 29, 3.15 pm, 2022. Seminar in Statistics.

Importance sampling as a mindset
Víctor Elvira, School of Mathematics, University of Edinburgh
Abstract: Importance sampling (IS) is an elegant, theoretically sound, flexible, and simple-to-understand methodology for the approximation of moments of distributions in Bayesian inference (and beyond). The only requirement is the point-wise evaluation of the targeted distribution. The basic mechanism of IS consists of (a) drawing samples from simple proposal densities, (b) weighting the samples by accounting for the mismatch between the targeted and the proposal densities, and (c) approximating the moments of interest with the weighted samples. The performance of IS methods directly depends on the choice of the proposal functions. For that reason, the proposals have to be updated and improved with iterations so that samples are generated in regions of interest. In this talk, we will first introduce the basics of IS and multiple IS (MIS), motivating the need of using several proposal densities. Then, the focus will be on motivating the use of adaptive IS (AIS) algorithms, describing an encompassing framework of recent methods in the current literature. Finally, we revisit particle filters and discuss alternative theoretical derivation of existing filters, explore the connections with multiple importance sampling, and propose a new family of methods with superior performance.
Location: John von Neumann (CHANGE OF ROOM!) .

Tuesday, December 6, 3.15 pm, 2022. Seminar in Statistics.

Spectral approaches to speed up Bayesian inference for large stationary time series data
Matias Quiroz, Department of Statistics, Stockholm University
Abstract: This talk will discuss some recent approaches to speed up MCMC for large stationary time series data via data subsampling. We discuss the Whittle log-likelihood for univariate time series and some properties that allow estimating the log-likelihood via data subsampling. We also consider an extension to multivariate time series via the multivariate Whittle log-likelihood and propose a novel model that parsimoniously models semi-long memory properties of multivariate time series.
Seminar slides
Location: John von Neumann (CHANGE OF ROOM!) .

Wednesday, December 7, 10.15 am, 2022. Seminar in Mathematical Statistics.

Multiple testing of mean values in multivariate data with BCS variance structure
Daniel Klein, Pavol Jozef Šafárik University, Slovakia
Abstract: common problem in multivariate data is that the number of unknown parameters in a model can be close or even higher than sample size. This may cause problems in statistical inference. Therefore, models with special variance structure are studied by many authors, where the number of covariance parameters is reduced by some restrictions on parameter space. One of possible structure could be so called block compound symmetry (BCS) structure. In this talk we will be studying estimation and testing such a structure as well as test of mean values assuming BCS structure. This is joint work with Ivan Žežula.
Location: Hopningspunkten.