Hide menu

The LiU Seminar Series in Statistics and Mathematical Statistics



Tuesday, March 8, 3.15 pm, 2022. Seminar in Statistics.

Two methods of conjoint summands of generating bivariate and trivariate normal pseudo-random numbers
Piotr Sulewski
, Institute of Exact and Technical Sciences, Pomeranian University
Abstract: The primary aim of the paper is to put forward methods of generating two dimensional (2D) and three dimensional (3D) normal pseudo-random numbers (NPRNs). The first method is based on uniformed random variables within (0,1), while the second method is based on normal random variables. Joint summands in these methods make resulting normal numbers correlated. The aim of the paper is to compare the existing methods of generating 2D and 3D NPRNs with new proposals. An in-depth analysis of the discussed methods allowed to compare not only the time of generating NPRNs but also to calculate the computational complexity. The obtained results indicate that one of the proposed methods has the best computational complexity and is faster than the others.
Location: Online via Zoom. Please email Krzysztof Bartoszek for invitation to Zoom meeting.

Tuesday, March 22, 3.15 pm, 2022. Seminar in Statistics.

Approximation of stochastic equations with irregular drifts via sewing
Konstantinos Dareiotis
, School of Mathematics, University of Leeds
Abstract: In this talk we will discuss about the rate of convergence of numerical schemes for stochastic equations with irregular drifts. Our approach relies on regularisation-by-noise techniques and more specifically, on the recently developed stochastic sewing lemma. The advantages of this approach are numerous and include the derivation of improved (optimal) rates and the treatment of non-Markovian settings. I will present results for equations in finite and infinite dimensions. This talk is based on joint works with Oleg Butkovsky, Khoa Lê, and Máté Gerencsér.
Location: Online via Zoom. Please email Krzysztof Bartoszek for invitation to Zoom meeting.

Tuesday, April 26, 3.15 pm, 2022. Seminar in Statistics.

The unresolved phylogenomic tree of butterflies and moths (Lepidoptera): assessing the potential causes and consequences
Niklas Wahlberg
, Department of Biology, Lund University
Abstract: The field of molecular phylogenetics is being revolutionized with next-generation sequencing technologies making it possible to sequence large numbers of genomes for non-model organisms ushering us into the era of phylogenomics. The current challenge is no longer how to get enough data, but rather how to analyse the data and how to assess the support for the inferred phylogeny. In this talk I will focus on one of the largest animal groups on the planet—butterflies and moths (order Lepidoptera), whose phylogeny remains unresolved despite several recent phylogenomic studies. In our study, we assess the potential causes and consequences of the conflicting phylogenetic hypotheses. With a dataset consisting of 331 protein-coding genes and the alignment length over 290,000 base pairs, including 200 taxa representing 83% of lepidopteran superfamilies, we compare phylogenetic hypotheses inferred from amino acid and nucleotide alignments. We find that especially the placement of the superfamily Gelechioidea is incongruent, which is likely due to compositional bias and possible other model violations. Furthermore, we employed a series of analyses to dissect our dataset and demonstrate that there is sufficient phylogenetic signal to resolve much—but not all—of the lepidopteran tree of life. The relationships among superfamilies within Ditrysia, the most species rich lepidopteran clade containing 98% of the extant species, remain poorly resolved. We conclude that taxon sampling remains an issue even in phylogenomic analyses and recommend that poorly sampled highly diverse groups, such as Gelechioidea in Lepidoptera, should receive extra attention in the future.
Location: Alan Turing

Tuesday, May 17, 3.15 pm, 2022. Seminar in Statistics.

Working as a biostatistician in the pharmaceutical industry at 4Pharma
Susanna Lövdahl
, 4Pharma
Abstract: 4Pharma is a consulting company with directions toward data collection and statistical analysis for clinical studies. The company was established in Finland 2022 and have a close collaboration with Helsingfors university who is the main owner. 4Pharma provides expert support to companies who developed medicines, medical devices and supplements to collect and analyze data that is the basis for regulatory approval around the world. Biostatisticians, Clinical Data Managers and Statistical Programmers work closely with physicians and research staff to ensure robust results. As a consultant, you work extensively with research in many different disease areas that can lead to a better future for patients.
Location: Alan Turing


Page responsible: Krzysztof Bartoszek
Last updated: 2022-08-29