We are privileged to have a little stream of distinguished but friendly guests, locals & foreigners, where we listen to and discuss some of their ongoing research themes, look out for common interests, etc.
In case you perhaps feel you might deserve to be on our guest list, contact us.
- Riccardo De Bin is the most recent recruit to the Statistics Division at the UiO Department of Mathematics, and was our guest 2 March 2017. Of course we already know him pretty well, and have enjoyed his two seminar talks here over the past few months, on aspects of his work. We interviewed him for further details both about his earlier work experiences (Padova, Pederobba, Northwestern, München, Nijmegen) and his ongoing research projects. These involve integrated likelihoods (as partly contrasted with likelihood profiling), increasing dimension for nuisance parameters, boosting, pairwise likelihoods, and applications in the biomedical directions of things. Our conversation led to certain ideas which will be followed up later. He earned his FocuStat cup.
- The statistics group at our department at UiO is privileged to have Peter Müller as a visitor this autumn. He's among the Famous Ones in Bayesian Nonparametrics (BNP) and also in medical statistics generally. He was our FocuStat guest 29 September 2016, and we interviewed him about his career moves (from Wien to Purdue to Duke to M.D. Anderson to University of Texas, Austin) and his work themes. Nils thinks the sequence of BNP conferences so far amounts to 1997 Belgirate, 1999 Reading, 2001 Ann Arbor, 2004 Roma, 2006 Jeju, 2007 Cambridge, 2009 Torino, 2011 Veracruz, 2013 Amsterdam, 2015 Raleigh (and 2017 Paris), and there's only person having been to all n of these, namely Peter (Nils has only been to n - 1 of them). Peter has written extensively about BNP, and has coauthored several books about this topic, fusing theory and applications. One of these is Bayesian Nonparametrics (Cambridge University Press, 2010), by Holmes, Hjort, Müller and Walker. Peter will also take part in FocuStat's Research Kitchen, October 2016.
Abhik Ghosh visited us 9 June 2016. He has a PhD on minimum divergence and minimum disparity estimators with Ayanendranath Basu at the Indian Statistical Insitute (for which he won a prize for the best PhD in Indian statistics and probability 2014-2015), and is currently a PostDoc with the Department of Biostatistics here at Oslo. We discussed aspects of both Abhik's earlier and ongoing work, touching also linear mixed models in higher dimension, with applications in epidemiology and perhaps genetics.
Kathrine Frey Frøslie was our guest 28 April 2016. She's now at least locally famous for her Statistrikk blog, featuring statistiches versions of everything from boxplots (over her own publication history) to birth-by-month distributions. She told us about current work concerning glucose curves, and we learned about bukspyttkjertelen and the liver and our inner bodies and how these may be monitored and mapped via functional data analysis models. Gudmund and Kristoffer might try their hands on certain prediction problems associated with such data.
Ingrid Van Keilegom (Institute de statistique, Université catholique de Louvain) visited us in September 2014, but was back in March 2016, both to work with Nils LH and to be an opponent in Martin Jullum's PhD disputation. Due to the terror attacks in Bruxelles her travel was delayed and made much more complicated, so the carefully laid plans involving Brahms and Schubert and Haydn with Céline and Nils did not come to fruition (this time around). She told us about current work of hers, including themes of empirical likelihood and cure models. She is also the recepient of an ERC Advanced Grant, with generous five-year funding for a project starting August this year.
- Daumantas Bloznelis, soon-to-finish PhD student at the School of Economics and Business at the Norwegian University of Life Sciences, Ås, visited us 11th of November 2015. Daumantas works with different complex models for the joint behaviour of multiple time series, leading e.g. to econometric short- and long-term predictions for the economics of the salmon industry. He's a polyglotted fellow who also can sing Händel and play Telemann's recorder sonatas and can beat us all in karate -- and he plans to widen his statistical horizon. We gave him the Peace Piece (with the splendid Užupis Constitution Song).
- Alan Gelfand, the Duke Professor of Statistics and Decision Sciences at Duke, was our specially invited "critical friend" for the week October 13-16, 2015. Alan is utterly famous on the statistical circuits, particularly well known for his work in Bayesian statistics and in time-space modelling and applications. His role for this week was to look friendly but critically into our various projects and offer suggestions for improvements. Alan forced Nils to wear the Alan Gelfand t-shirt, appropriately chained in Monte Carlo Markov chains.
- Monica Musio, Università degli Studi di Cagliari Matematica ed Informatica, and Fabian Krüger from the Heidelberg Institute of Theoretical Studies, visited us in September 2015, in connection with our Research Kitchen on proper scoring rules, minimum divergence estimation, robust model selection, spatial modelling, etc. Also Thordis Thorarinsdottir from the Norwegian Computing Centre took active part in these discussions. We also had occasion to discuss various other matters!
- Peder Østbye visited us 10th of April 2015, giving a presentation of his work on focused information criteria in situations where the actual cost of covariates play a role for the model selection; applications arise in fields as diverse as business law and biostatistics. Other issues came up in the discussion, including thoughts on further work along such lines. The FocuStat group plans to host an international-level workshop on "focused model selection and model averaging" around May 2016, and then Peder ought to be one of the speakers.
- Ingrid Glad and Arnoldo Frigessi visited our group 20th of March 2015, where we discussed ongoing projects and research directions for both FocuStat and the new Centre for Innovation Big Insight. There will be opportunities down the road for so-called mutual benefits. One scenario is that Nils LH can be supervisor or co-supervisor for one more PhD and one more PostDoc, funded by Big Insight, but with sufficient overlap with FocuStat in terms of methodology and applications.
- Per Mykland paid us a visit 13th of October 2014, in connection with his spending a week in Oslo to meet both UiO colleagues and talk to the finance people at Norges Bank. We learned about aspects of his own ongoing research work, and enjoyed his very useful comments and responses to our group's ongoing projects and plans for 2015.
- Kristoffer Hellton visited our group 6th of October, giving us an occasion to discuss work he is currently engaged in and wishes to pursue further, partly inspired by models and estimation issues encountered with the Measurement Error Modelling group.
- Ingrid Van Keilegom (Institute de statistique, Université catholique de Louvain), September 2014: Ingrid, Ian and Nils have worked together on certain extensions of empirical llikelihood theory, and are now pursuing yet further related themes. For their research kitchen visit with FocuStat they also started discussing certain time-to-cure models related to gamma process crossing times.
- Ian McKeague (Department of Biostatistics, Columbia University, New York), September 2014: See above.
- Jana Timková (Department of Statistics, Charles University, Prag), May 2014: She works with Nils on various aspects related to the monotone version of Aalen's linear hazard regression model for survival data, which entails studies of both frequentist and nonparametric Bayesian procedures.
- Daniel Roy (now with the Department of Statistical Sciences, University of Toronto), March 2014: He works with certain probabilistic generalisations of Nils' Beta Processes, called the continuum of urns process, with potential applications in both machine learning and in statistics.