These are talks that I’ve held on various occasions.

• Microbiome data: what’s in the bag? When you do a typical microbiome study you get measurements constrained to sum to a certain number (called compositional data). Is this useful? Can we cheat our way toward non-compositional measurements? Presented at CANS day, 2023. [slides]
• DAGs: tools for reasoning about causes. A guest lecture I’ve given about once a semester for a couple of years on the ue of DAGs in causal inference. Most recently for the Department of Community Medicine, UiT, October, 2023. [slides]
• Introdution to Bayesian Inference. Guest lecture given to the advanced epidemiology course, Dep’t of Community Medicine, UiT. October, 2021. [slides]
• Version control, &c. A small workshop on version control for the Dep’t of Community Medicine, UiT. November 25th, 2020. [slides]
• Opposing forces in big data. On expectation propagation. Presented at the HDL lab meeting, UiT. October, 2020. [slides]
• One weird trick to quantify uncertainty One hour workshop teaching the bootstrap to our lab. [slides] [rstudio.cloud project]
• Scientific cartography A talk about causal DAGs for Norwegian Women and Cancer’s systems epidemiology. Given by me and Karina Standahl Olsen in November, 2019. [slides]
• Learn to live with lack of rejection A talk for Norwegian Women and Cancer’s systems epidemiology group by myself and Kajsa Møllersen. [slides]
• Small data: practical modeling issues in human-model -omic data Defense for the degree of Ph. D. 8th of February, 2019. [slides]
• Variable selection in genomics — Methods, challenges, and possibilities Trial lecture for the degree of Ph. D. 8th of February, 2019. [slides]
• Shrinkage estimation for fun and profit BDPS lab meeting, University of Tromsø. 19th of February, 2018. [slides]
• Predicting breast cancer metastasis from blood samples — “On variance and other problems” Presented at the Breast Cancer and the Immune System workshop, Barbados, 2017. [slides]
• Predicting breast cancer metastasis from blood gene expression using time-to-diagnosis data to derive signature estimates. Poster presented at the 2017 Statistical Methods for Postgenomic Data workshop in London. [poster]
• The pragmatic R programmer. Intro to writing reusable R code and bundling it in packages. Given at the Department of Community Medicine, University of Tromsø. 17th of November, 2016. [slides] [demo package]
• Version control and you. Intro to version control software (Git) and its place in reproducible data analysis. Given at the Department of Community Medicine, University of Tromsø. 16th of November, 2016. [slides]
• Predicting cancer progression from blood gene expression (Alternate title: Some negative results). Talk given for the Machine Learning @ UiT weekly seminar 23rd of October, 2015.