Tag Archives: science

Workshops

I had a very full week last week, with the annual Bayes on the Beach (BOB) at the Gold Coast (Mon-Wed) and Bayesian Optimal Design of Experiments  (BODE) on Friday.

BOB is an annual workshop/retreat, run by Kerrie Mengersen and the BRAG group at QUT, that brings together a bunch of Australian and international statisticians for a few days of workshops, tutorials, presentations and fun in the sun. This year was, I think, my fourth year at BOB.

One of the recurring features is the workshop sessions, where around three researchers each pose a problem to the group and everyone decides which one they’re going to work on. This year I was asked to present a problem based on the air quality research I do and so my little group worked on the issue of how to build a predictive model of indoor PM10 based on meteorology, outdoor PM10 and temporal information. We were fortunate to have Di Cook in our group, who did a lot of interesting visual analysis of the data (she later presented a tutorial on how to use ggplot and R Markdown). We ended up discussing why tree models may not be such a great idea, the difference in autocorrelation and the usefulness of distributed lag models. It gave me a lot to think about and I hope that everyone found it as valuable as I did.

The two other workshop groups worked on ranking the papers of Professor Richard Boys (one of the keynote speakers) and building a Bayesian Network model of PhD completion time. Both groups were better attended than mine, which I put down to the idea that those two were “fun” workshops and mine sounded a lot like work. Still, a diverse range of workshops means something for everyone.

James McGree (QUT) asked me if I could come to the BODE workshop to discuss some open challenges in air quality research with regards to experimental design. I gave a brief overview of regulatory monitoring, the UPTECH project’s random spatial selection and then brought in the idea that the introduction of low cost sensors gives us the opportunity to measure in so many places at once but we still need to sort out where we want to measure if we want to characterise human exposure to air pollution. While it was a small group I did get to have a good chat with the attendees about some possible ways forward. It was also good to see Julian Caley (AIMS) talk about monitoring on the Great Barrier Reef, Professor Tony Pettitt (QUT) talk about sampling for intractable likelihoods and Tristan Perez (QUT) discuss the interplay between experimental design and the use of robots.

It’s been a great end to the year to spend it in the company of statisticians working on all sorts of interesting problems. While I do enjoy my air quality work and R usage is increasing at ILAQH it’s an entirely different culture to being around people who spend their time working out whether they’re better off with data.table and reshape2 or dplyr and tidyr.

Advertisements

Australia-China Centre turns 1

Has it already been a year?

This week the Australia-China Centre for Air Quality Science and Management had its second annual meeting, at QUT. We got updates on the various research activities that have happened, are happening and are planned. There’s lots of interesting stuff being done to tackle a variety of problems, such as reducing workplace exposure to air pollution, quantifying the exposure of individuals and using unmanned aerial vehicles to measure air quality.

IMG_1215.jpg

Tuesday night we had the conference dinner out at the Mount Coot-tha Botanic Gardens, at the function space at the cafe/restaurant out there. I don’t think I’ve been there since my cousin’s wedding reception 15-20 years ago. I really liked that efforts were made to ensure each table had a mix of senior professors, mid- and early-career researchers and PhD students. It made for a very inclusive dinner and many different topics of conversation. Luckily I was sat with a co-worker with whom I could trade my fish entree and mains for something a little more land-based. There was even a birthday cake (chocolate mousse cake) and a number of people joined in singing “Happy Birthday” to the ACC.

Wednesday we spent the day workshopping the various planned projects to determine what issues need to be addressed in the collection and analysis of data. I ended up sitting with a group looking at the impacts of indoor temperature on mortality rates, particularly trying to estimate the relative risk of extreme heat and cold. It was good to be confronted with some new challenges to think about, rather than the same stuff I’ve been working on almost non-stop this year.

All in all, it was a good meeting even though the stress levels around here were through the roof in the lead-up. I ended up taking photos of nearly all of the presenters on the Tuesday as well as group photos with our Chinese collaborators and special invited guests.

Just got back from China

I’ve spent the last few days travelling to and from Beijing, China for the launch of the new Australia-China Centre for Air Quality Science and Management. This is a huge thing for us at ILAQH because it sets up an international collaborative agreement between a bunch of institutions in Australia and China who each have a different set of expertise that they can bring to the table, allowing us to undertake more ambitious projects than before and seek funding from a wider range of sources.

There was a lot of prep and behind the scenes meetings to take place on the first day, so Mandana (a colleague of mine) and I took a trip downtown to the Forbidden City and Tian Anmen Square and did some shopping Wangfujin. It would have been cold enough without the wind but it was such a clear day and everything was wonderful.

Day 14 - Tian Anmen Square Beijing

The first day of the launch, held at CRAES, featured the constituent groups giving a short presentation about their work. There’s a lot of great people working on some really interesting stuff and I’m very excited about the prospect of working with some of them over the coming years.

Day 15 - Lina

The second day had us split into three groups to propose various objectives and projects and put names on paper for who might be good leaders or key players in these fields as part of our centre. I joined the Transport Emissions group to discuss the control of emissions at their source, the investigation of atmospheric transformation processes and the development and uptake of new technologies. A lot of the ground work had already been laid at a planning meeting earlier in the year but it was good to put together some more concrete research topics.

After this, we went out to the National Jade Hotel for dinner, where we got to try another of the varied styles of Chinese cuisine; this time from a coastal region in North-Eastern China. I wish I had’ve paid more attention to the names of the various styles, but I enjoyed trying everything over the course of the trip, even the tripe.

Day 16 - The most important decision

The final day saw us tidying up the proposals, and an early finish meant that I got the afternoon off with Mandana, Felipe and Dion. After stumbling our way through a menu with pictures but no English translations, we had a big lunch and set off on the subway to the Temple of Heaven. It was certainly warmer than the Forbidden City (less stone, more trees) and was a very peaceful and pleasant end to the trip as we sat down at a bakery café and discussed If You Are The One while eating cream buns and drinking coffee (or in my case, peach black tea with milk). After heading back to the hotel, Mandana and I took a walk around the Bird’s Nest stadium which was only a block from our hotel. It looks like Beijing is putting effort into maintaining the area as a public plaza rather than just the grounds of a sports stadium, so even late in the cold evening it was full of families and groups of friends walking, talking and taking photographs.

Day 17 - Bird's Nest

An early morning taxi to the airport saw the start of 18 hours of travel. It’s nice to be back in one’s own bed, but I’m off to the airport again tomorrow for a 5:30am flight to Sydney for a workshop on exposure assessment with colleagues from the Centre for Air quality and health Research and evaluation. After a week of disastrously bad coffee, I’m glad that I booked accommodation which advertises itself as being 15m from the Toby’s Estate café.

The problem with p values

A coworker sent me this article about alternatives to the default 0.05 p value in hypothesis testing as a way to improve the corpus of published articles so that we can actually expect reproducability and have a bit more faith that these results are meaningful. The article is based on a paper published in the Proceedings of the National Academy of Sciences which talks about mapping Bayes Factors to p values for hypothesis tests so that there’s a way to think about the strength of the evidence.

The more I do and teach statistics the more I detest frequentist hypothesis testing (including whether a regression coefficient is zero) as a means of describing whether or not something plays a “significant” role in explaining some physical phenomenon. In fact, the entire idea of statistical significance sits ill with me because the way we tend to view it is that 0.051 is not significant and 0.049 is significant, even though there’s only a very small difference between the two. I guess if you’re dealing with cutoffs you’ve got to put the cutoff somewhere, but turning something which by its very nature deals with uncertainty into a set of rigid rules about what’s significant and what’s not seems pretty stupid.

My distaste for frequentist methods means that even for simple linear regressions I’ll fire up JAGS in R and fit a Bayesian model because I fundamentally disagree with the idea of an unknown but fixed true parameter. Further to this, the nuances of p values being distributed uniformly under the Null hypothesis means that we can very quickly make incorrect statements.

I agree with the author of the article that shifting hypothesis testing p value goal posts won’t achieve what we want and I’ll have a bit closer a read of the paper. For the time being, I’ll continue to just mull this over and grumble when people say “statistically significant” without any reference to a significance level.

NB: this post has been in an unfinished state since last November, when the paper started getting media coverage.

Revising another paper

We got a paper back from the reviewers a few days ago and there are some comments requesting revisions to the explanation of the statistical methods and the analysis. It’s interesting coming back to this paper, about a year after I last saw it (it’s been sent around to a few different journals to try to find a home for it). The PhD student who is the main author got into R and ggplot2 last year and has done some good work with linear mixed effects models and visualisation but some of the plots are the same sort of thing one might do in Excel (lots of boxplots next to each other rather than making use of small multiples).

So now I get to delve back into some data and analysis that’s about a year old with fresh eyes. Having done more with ggplot2 over the last 12 months, there are some things that I will definitely change about this. The student and I had a chat this morning about how to tackle it, and we’re trying to choose the best way to split up our data for visualisation: 6 treatments, 4 measurement blocks, two different measures (PM2.5 mass concentration and PNC), a total of 48 boxplots, density plots or histograms.

A paper with another PhD student has had its open discussion finalised now, which means more writing is probably going to happen. I find ACP‘s model quite interesting. The articles are peer reviewed, published for discussion, and then revised by the authors for final publication. I guess it spreads the review work out a bit and allows for multiple voices to be heard before final publication, each with a different approach and background.

That feeling when former students contact you

Last year I had a student in SEB113 who came in to the subject with a distaste for mathematics and statistics; they struggled with both the statistical concepts and the use of R throughout the semester and looked as though they would rather be anywhere else during the collaborative workshops. This student made it to every lecture and workshop though and came to enjoy the work of using R for statistical analysis of data; and earned a 7 in the unit.

I just got an email from them asking for a reference for their VRES (Vacation Research Experience Scheme) project application. Not only am I proud of this student for working their butt off to get a 7 in a subject they disliked but came to find interesting, but I am over the moon to hear that they are interested in undertaking scientific field research. This student mentions how my “passion for teaching completely transformed my (their) view of statistics”, and their passion for the research topic is reflected in the email.

This sort of stuff is probably the most rewarding aspect of lecturing.

Lotka-Volterra and Bayesian statistics and teaching

One of the standard population dynamics models that I learned in my undergrad mathematical modelling units was the Lotka-Volterra equations. These represent a very simple set of assumptions about populations, and while they don’t necessarily give physically realistic population trajectories they’re an interesting introduction to the idea that differential equations systems don’t necessarily have an explicit solution.

The assumptions are essentially: prey grow exponentially in the absence of predators, predation happens at a rate proportional to the product of the predator and prey populations, birth of predators is dependent on the product of predator and prey populations, predators die off exponentially in the absence of prey. In SEB113 we cover non-linear regressions, the mathematical models that lead to them, and then show that mathematical models don’t always yield a nice function. We look at equilibrium solutions and then show that we orbit around it rather than tending towards (or away from) it. We also look at what happens to the trajectories as we change the relative size of the rate parameters.

Last time we did the topic, I posted about using the logistic growth model for our Problem Solving Task and it was pointed out to me that the model has a closed form solution, so we don’t explicitly need to use a numerical solution method. This time around I’ve been playing with using Euler’s method inside JAGS to fit the Lotka-Volterra system to some simulated data from sinusoidal functions (with the same period). I’ve put a bit more effort into the predictive side of the model, though. After obtaining posterior distributions for the parameters (and initial values) I generate simulations with lsode in R, where the parameter values are sampled from the posteriors. The figure below shows the median and 95% CI for the posterior predictive populations as well as points showing the simulated data.

lvThe predictions get more variable as time goes on, as the uncertainty in the parameter values changes the period of the cycles that the Lotka-Volterra system exhibits. This reminds me of a chat I was having with a statistics PhD student earlier this week about sensitivity of models to data. The student’s context is clustering of data using overfitted mixtures, but I ended up digressing and talking about Edward Lorenz’s discovery of chaos theory through a meteorological model that was very sensitive to small changes in parameter values. The variability in the parameter values in the posterior give rise to the same behaviour, as both Lorenz’s work and my little example in JAGS involve variation in input values for deterministic modelling. Mine was deliberate, though, so isn’t as exciting or groundbreaking a discovery as Lorenz but we both come to the same conclusion: forecasting is of limited use when your model is sensitive to small variations in parameters. As time goes on, my credible intervals will likely end up being centred on the equilibrium solution and the uncertainty in the period of the solution (due to changing coefficient ratios) will result in very wide credible intervals.

It’s been a fun little experiment again, and I’m getting more and more interested in combining statistics and differential equations, as it’s a blend of pretty much all of my prior study. The next step would be to use something like MATLAB with a custom Gibbs/Metropolis-Hastings scheme to bring in more of the computational mathematics I took. It’d be interesting to see if there’s space for this sort of modelling in the Mathematical Sciences School’s teaching programs as it combines some topics that aren’t typically taught together. I’ve heard murmurings of further computational statistics classes but haven’t been involved with any planning.