ISI Durban
The ISI 2009 conference was held lst week in Durban, South Africa. I quite enjoyed the previous one in Lisbon and was invited to this one so I thought I would go back for seconds. Below the fold are some impressions (considered an uncontroversial as ever) on the academic and non-academic aspects of the meeting.
First, I will admit that I am cranky with the organizers for listing my special session wrongly on the timetable and then not advertising the reschedule, or the location. I couldn’t even find the room myself. Consequently there was an audience of the six which included the four speakers. Thanks are due to Paul Kabaila for making up half of the non-presenting audience.
The more serious gripe with the conference was personal safety. In Durban you cannot actually walk the streets. Anyone who might have money (which includes all whites) is fairly likely to get mugged and robbed. The hotels were no more than 400 meters from the conference venue, the roads were open and wide, the sun was shining. But the conference organizers and hotel managers insisted that we could not walk it, even at 9 in the morning and even in large groups. After dark, don’t even think about it. Unaccountably, there is very little visible police presence. It is no way to live and, I would have thought, death to the South African economy.
Probably the non-academic highlight of the conference was the game park tour. You are actually safer amongst the lions and rhinos than you are walking the predatory streets of Durban! It was interesting to see the animals interacting together in a quasi natural environment rather than being on display in separate cages. For instance, the smaller animals all hang around the giraffes and zebras who are the best as seeing and smelling predators respectively. Free riding is rife in nature.
Turning to the substance of the meeting, I don’t know what your experiences are, but I am finding more and more that, at large conferences at least, most of the talks are simply awful. There are entire sessions of people reporting uninteresting applied work, rediscoveries of well known results, papers that are technically incomprehensible to anyone, speakers reading directly from slides in unintelligible English and an increasing incidence of “no shows”.
Unlike some other disciplines, when you send your abstract to a statistics conference and receive the email saying “your paper has been accepted”, we all know that nobody has actually read the paper*. There seems to be no quality control at all as far as I can see. I guess the organizers just want as many paying participants as possible to cover their costs.
But enough of the negative! There were a couple of sessions I enjoyed where I learned something new and fun.
The first concerned distributions theory, specifically distributions that are defined more easily through their quantile function (i.e. inverse cdf) than their cdf or pdf. Examples are the Davies distribution and the generalized lambda distribution. Ultimately, you can always get the cdf numerically and just use standard methods but the alternative quantile view leads to new ideas and techniques. First, there are a bunch of moments from integrating the quantile function with respect to Legendre polynomials - called Legendre or L moments. The estimators are linear combinations of order statistics. If the first Legendre moment exists then they all exist. The distribution with standard zero Legendre moments is the uniform distribution, not the normal. These exotic moments generate alternative method of moments estimators. Second, quantile functions combine simply. Location and scale transforms of the quantile function transform the location and scale of the variable. Also a linear combination of quantile functions is another quantile function but it does not correspond to a mixture. Only the odd moments are affected. This makes it easy to augment standard distributions into larger families. Anyway, while there was nothing earth shattering and probably nothing presented that was unknown, it was a rather new and fresh – to me at least.
The second session I enjoyed was on adversarial risk analysis by which they mean quantitative strategies against terrorists. David Banks, the organizer and a main speaker of the session began by personally introducing himself to each audience member (there were only about 10 of us), thanking us for coming and asking us what our interests and background were. I use this angle for cutting the ice with a new MBA class but I have never seen it used at a statistical conference before.
The session was about using decision trees where there are two players with different utilities on the possible outcomes and where the utilities and probabilities for the terrorist are unknown. Standard game theory won’t work because the utilities are not the same for the two players nor known to each other. Standard decision analysis also fails in practice because the unknown actions of the adversary are not meaningfully dealt with using probability.
The proposed approach was put yourself in the terrorist shoes, to elicit a distribution for the terrorist’s assessment of your utilities and probabilities, and use this to generate a distribution for his probabilities. This can also be pushed one step further back, where you make estimates of what the terrorist will likely assume about your probabilities and utilities. Pushing the subjective probabilities deeper can actually lead to better robustness, kind of like Bayesian hyper priors having less influence than direct priors. So it is a combination of game theory, subjective probability elicitation and decision trees. Most of the details are in this recent JASA paper.
* Note to self. I might test this assertion by sending an abstract containing deliberate nonsense and see if it is accepted – like a nerdy statistical version of Angry Penguins.
You can follow any responses to this entry through the RSS 2.0 feed. You can leave a response, or trackback from your own site.
September 2nd, 2009 at 10:33 am
On the non-statisical side of this review, I don’t agree that security was as dire as suggested. I had walked to and from the conference centre from my hotel on the beach front and did not have any problems. This doesn’t mean that I think it was unsafe - I wouldn’t have attempted it in the dark - but it wasn’t as bad as suggested. Security people I talked to at the conference centre were pretty relaxed about security, provided you were cautious.
It’s probably fair to say that South Africa is spending an awful lot of effort ensuring that they can deal with the safety of visitors for major events, such as this conference and the forthcoming World Cup 2010. I think they did a pretty good job, in a city that can sometimes be quite dangerous. But, I’ll admit, one person’s perception of security risk is different from another’s perception.
The highlight session for me was the Presidential Invited Session with Iain Chalmers (”The evolution of controlled intervention trials…”), Mark Orkin (roughly, observations on survey statistics in South Africa) , and Trevor Hastie (”Modern trends in data mining”). Well, most entertaining, anyway…
September 2nd, 2009 at 10:35 am
Sorry your comment and others got trapped in Spam. My perceptions about security were almost entirely based on advice from the conference organisers and the hotels and restaurants. They were uniformyl insistent that I should not walk anywhere.
I agree that it looked safe walking from the beachfront to the conference centre, but the locals assured me that it wasn’t. So you were lucky I guess.
September 2nd, 2009 at 10:36 am
There are aspects of ISI meetings which are planned to be different to most other statistics meetings. There is a big emphasis on diversity which (naturally) means larger variance, but may be more likely to produce new insights. This is not well-publicised. A few examples:
1. A major objective is to promote communication between specialists in different fields of statistics - eg survey statisticians & mathematical statisticians, those interested in methodology & those interested in a particular field of application (eg environment). For example a session on the role of statisticians in climate change involved a mathematical statistician, an official statistician and someone from a commercial background.
2. About 1/3 of the sessions are invited paper sessions and the expectation is that the participants will not be a ‘comfortable’ group of colleagues from one country - the speakers should be from around the world.
3. There is a special emphasis on involving statisticians from developing countries (and, for 2011 at least, female statisticians and early career statisticians).
4. The aim is to make each Invited Paper session topical and of high quality.
The role of the invited discussant is also more widespread at ISI than in most other large meetings. For 2011 we will be seeking to make the discussion a more important part of the sessions (I’m chair of the Programme Co-ordinating Committee for 2011.)
With regard to contributed papers and quality control, anyone entitled to be at the meeting is entitled to present a contributed paper.
I agree that the security was incredibly frustrating (and concerning). The benefit of having the meeting in South Africa is that it gave enormous boost to statisticians in Africa - I think there were about 500 at the meeting and a specific goal was to build capacity in Africa, in statisticians and in schools. This seems to have been achieved.
And no review of the meeting could ignore the high quality of the African music at the opening ceremony - a high point for me.
September 2nd, 2009 at 10:41 am
Sorry also to you Murray. I will turn the second spam filter off. I understand that anybody is allowed to present a paper at a stats conference. Two points I would make. In other disciplines you have to get over a hurdle to present. Second, the ISI sends an email that your paper has been “accepted” which, to me at least, suggests some quality control process.
In other disciplines again, conference papers are briefly referees and sorted into two streams, with suitably euphemistic labels. The lower stream is very poor quality – but at least useful for students and their supervisors. I would suggest that we should do the same at statistics conferences. Invited paper sessions are usually OK, but that still leaves a lot of sessions where I have little idea in advance whether or not I am wasting my time.
September 17th, 2009 at 9:48 pm
Chris is quite correct in his comments on personal safety and stating “It is no way to live” – my Argentine wife reminds me of this on a daily basis (not that I need the daily reminder…). I could write essays on the crime and safety problems …hmm… crime and safety challenges to be politically correct, but unfortunately my views will NOT be politically correct and are hence best kept in my head.
Regarding the ISI conference and its sessions: I was the organizer of the first session (an STCPM) mentioned by Chris and want to thank him for his kind words. Together with a student of mine, I furthermore presented the paper on quantile functions and L-moments in this STCPM. Just to clarify, the L in L-moments does not stand for “Legendre”. Quoting Hosking (1990), “L-moments are expectations of certain linear combinations of order statistics.”, so the L stands for “Linear”. There are 3 ways to define L-moments, and since we were promoting the use of quantile function modelling, we only gave the one definition in terms of the quantile function and the Legendre polynomials. I apologize that my student did not explain this better in our presentation (always nice when you can blame the co-presenter!). We are currently busy writing a paper on our ideas and methods – we are for example trying to find solutions to the question posed to me by Chris after the presentation – and will hopefully have it published soon.
Regarding Murray’s comments: I agree with his four points (examples). Points 1, 2 & 4 were definitely true for the ISI conference in Lisboa (my maiden ISI). Point 3 was one of the aims for the ISI conference in Durban. I cannot comment on behalf of other South Africans whether they benefited on having this prestigious conference at home. I certainly did. It is extremely expensive for researchers from South Africa (and sub-Saharan Africa) – especially those at the beginning of their research careers – to attend and participate in international conferences in the rest of the world. So naturally I tried to use this wonderful opportunity as best I could.
When I started organizing my STCPM, I was told in no uncertain terms that my session’s participants could not be just from South Africa, but should be from around the world. Fortunately my session conformed to this requirement with Australian, Turkish, and Polish participants. But it became soon clear to me at the conference that many of the other STCPMs did not follow this “rule”. Not that I personally have a big gripe regarding this. Inviting renowned researchers from around the world to participate in my STCPM has provided me with a lot of stimulus for my own research and doctoral studies. I have met wonderful people at the ISI with whom I hope to collaborate in future research.
What was a big disappointment to me at the ISI conference, was the (lack of) organization of the sessions and programme:
1. As with the other publications, Information Bulletin No. 3 contained beautiful photos and artwork. It did not contain my name in the “Index by Authors” (my surname starts with v…). Much worse, despite numerous e-mails from me to the organizers regarding the participants and order of presentation in my STCPM, the programme in the Bulletin was incomplete and completely wrong. To my surprise a lady from Italy, who had to withdrew her participation from my STCPM due to lack of funding, was listed to speak in another STCPM. Now I repeatedly communicated to the organizers that she would not be attending the ISI conference. So of course she was one of the “no shows” at the other STCPM (out of the 6 presenters listed in that STCPM, only 2 arrived, so together with the chair and me, we were 4 people…). Returning to the published programme, there was no coherency in the layout. For some sessions only the first author of each paper was given, while for others full names (that is, all names + the surname) of all authors of each paper were provided.
2. The 499-page Book of Abstracts is a waste of paper and I am sure that many copies ended up in dustbins (recycle bins?) in Durban. Why include page numbers if there is no index? Many abstracts of the “no shows” are in there, while quite a few presenters told me that their abstracts are “no shows” in the Book of Abstracts.
3. I did not receive the ISI 2009 Proceedings at the conference. Later I heard that the CDs were distributed at the closing ceremony. After politely asking the organizers for a copy, I last week received one. The Proceedings is as incomplete as the Book of Abstracts, if not more so. The Proceedings should contain all the papers and not just abstracts or a mixture of abstracts and papers.
To conclude, a conference is not only remembered by the quality of the coffee served or the success of the social programme. For those who could not attend a conference for whatever reason, the Programme, the Book of Abstracts and the Proceedings provide the only view of the conference. What a pity that the view for ISI 2009 will be blurry.
I do hope that Chris will go for thirds at ISI 2011 in Dublin. I am certainly planning to do so. And I am not expecting any African music at the opening ceremony…