Harley Wood Lecture Online

On the 6th of July, I had the great privilege of presenting the Astronomical Society of Australia’s (ASA) Harley Wood Lecture: “Life in a Finely Tuned Cosmos”. The Harley Wood Lecture was inaugurated in 1984 as an annual lecture in honour of the first President of the ASA. Harley Wood was Director of Sydney Observatory for over thirty years from 1943 to 1974. It was great to see nearly 300 people come out on a rainy Wednesday night.

Audio from the talk is now online at the Sydney Ideas Soundcloud. It also includes a response from Prof. Mark Colyvan, a philosopher at the University of Sydney, and a Q&A session. I don’t think that a video exists, but my slides are also online here if you want to follow along. Mark’s slides are here.

Read Full Post »

A Co-Blogger!

It’s probably a good time to remind ourselves that, in spite of the evidence of the past few years, this is a collaborative blog. I hope you enjoyed the glorious return of Berian (who named this blog, incidentally). I’m hoping he’ll write more about the overlap of astronomy and data science, and the difference between the academic and business world more generally.

Incidentally, Brendon Brewer is blogging some Bayesian thoughts over at Plausibility Theory, so please follow that blog. There’s some particularly interesting stuff on probabilities and thermodynamics. Matt Francis, after working for a time in Italy and at the Bureau of Meteorology, has also wandered into the business world. Matt was kind enough to come back to the astronomers at the University of Sydney and share his experiences outside academia.

Read Full Post »

Notes from Data Science Summit: ML in Production

I’m (Berian, that is, rather than Luke) attending the SF Data Science Summit today and tomorrow. I’m taking some rough notes as I go and want to publish them in digestible bits. One of the speakers I most enjoyed today was Carlos Guestrin (@guestrin), who gave a keynote and then a little 25-minute appendix later in the day. Here’s what I wrote down.

Read Full Post »

Conference Posters 101

The 2016 Astronomical Society of Australia’s Annual Scientific Meeting is over for another year – congratulations to all involved for a wonderful conference.

Naturally enough, I’ve been pondering conference posters for the last week. Here is my unsolicited advice.

Title

I’ll probably read a poster in spite of a boring title, but do yourself a favour and make it interesting. In particular, make your title a scientific claim. I got this idea from Josh Peek, who makes the title of every slide of his talks a summary of the slide, not just the theme. E.g. instead of “Observations”, put “We observed 234 galaxies”. Instead of “Modelling”, put “A warped disk fits our data”.

My preference (comment if you disagree) is for titles that are claims. E.g. Instead of “The GALAH Survey: An Overview”, put “GALAH will survey 1 million stars”.

Also, consider a subtitle: a one sentence summary of your poster.

Abstract

Always start the main body of text with an abstract or summary. I won’t read every word on every poster, so don’t make the point of your poster hard to find.

Also, put the introduction after the abstract. I found myself being irritated by posters with vague titles that then started with “There are lots of stars in the sky. Stars are bright. …” I found myself scanning through the poster to try to find what was new, what I could learn.

Language and Layout

• This is a lesson from writing grant proposals, but beware of soft verbs like characterise and classify. Even probe can be a bit weak. Tell us what the science goal is, even if you’re not there yet.
• Flow charts and diagrams are very useful. Equations are particularly useless on a poster.
• Light text on a dark background, or vice versa. Contrast,contrast,contrast.
• Don’t make the text too large. Readable from a metre away, not from across the room.
• Include a photo of yourself
• Labels, comments and arrows on plots are great. In particular, summarise the point of the plot in one sentence. E.g. “This model (arrow to line), which includes stellar feedback, best fits the data.”

Have I missed anything? Comment below!

Read Full Post »