Mike Irvine

Forgive me, for I am about to Bayes. Lesson: Don't trust intuition, for even simple prior+likelihood scenarios defy it. Four examples below, each producing radically different posteriors. Can you guess what each does? Revealed in next tweet >>

@TAH_Sci Would a good Bayesian be using an improper prior?

Our new study modelling the urgent need for #naloxone in the US. Using a #Bayesian model we developed for BC we estimated the probability of its use during a witnessed overdose. naloxoneneededtosave.org provides estimates for each state @DanCoombsUBC @TraciCGreen1

@WarwickComplex @JonathanSkipp Congratulations @JonathanSkipp !

So very proud of @JonathanSkipp who smashed the living heck out of their viva on Wednesday! Well done, Jon, we salute you! 👏👏👏 #phdone2021 #universityofwarwick

@JoeBHilton Obligatory XKCD xkcd.com/793/

The real reason to go Bayesian.

@mathdona @GaryMirams @UoNMaths Congratulations! 🎉🎉🎉

Oh! I forgot to post this on Monday: I'm officially officially a postdoctoral researcher at the School of Mathematical Science of University of Nottingham :) @GaryMirams @UoNMaths

Excellent discussion with @tedfieldglobal from @GlobalBC on our @ubcmedicine study – visits during the COVID19 peak pandemic @BCCHresearch globalnews.ca/news/7445146/b…

@dbhaskar92 @HeikoEnderling @ara_anderson @mkjolly15 @mathonco @ESmb2020 @MathDeptOrg Nice work! Anywhere we can view the poster online?

I got a poster prize at #SMB2020! Thanks to @HeikoEnderling @ara_anderson @mkjolly15 and all the judges in the @mathonco subgroup. Thank you @ESmb2020 organizers and @MathDeptOrg team for a wonderful experience! See you next year, hopefully in-person.

Academic lemma: the number of requests and emails will go towards infinity as the date goes towards when you're supposed to go on holiday.

Our team @CDCofBC & @ubcspph in #Vancouver is recruiting a quantitative #epidemiologist with focus on #COVID-19 and a #postdoc infectious disease #modelling to work with COVID-19 response team @CDCofBC. #epitwitter. DM me if interested. #BigData #YVRjobs @UBCmedicine

Long time frames to detect the impact of changing COVID-19 control measures. Tag line: it takes longer than 2 weeks to see the impact of a change in physical distancing (or other NPIs) with @MostlyStatic @CarolineColijn medrxiv.org/content/10.110…

@DanCoombsUBC Skype at least does have a feature for subtitle, not sure if the same is true of zoom?

I was reminded that many people rely on lip-reading in their daily lives and this is seriously hindered by masks. And makes a life of zoom/skype meetings tough.

Pleased our work on the use of mixture density networks for emulation in epidemiology has recently been published in @PLOSCompBiol. Check it out: doi.org/10.1371/journa… @MikeAIrvine @DeirdreHoll @QuentinCAUDRON

Yesterday's talk for those who missed it, or don't have Netflix: Real-time modelling of the 2020 coronavirus epidemic. youtu.be/dLVLEZMzIOk via @YouTube. Shoutout to @UBC cosmologist Douglas Scott for the invitation to speak in the physics colloquium.

@TAH_Sci Thanks for sharing Thomas. I got inspired by this (and others) to create one myself: sempwn.github.io/blog/2020/03/1…

I created a blog post that allows you to explore how interventions might impact the covid-19 pandemic sempwn.github.io/blog/2020/03/1…

Lots of talk at the moment surrounding the R0. I recently created an app to explore the concept for educational purposes only: sempwn.github.io/vaccine-dynami…

@QuentinCAUDRON Pacific Northwest in the fall though, would be better to predict when it isn't raining...