A new position for the of Professor Of Statistics and Data Science / Director of the [newly created] Warwick Data Science Institute has been posted. To quote from the job description, “the position arises from the Department of Statistics’ commitment, in collaboration with the Warwick Mathematics Institute and the Department of Computer Science, to a coherent methodological approach to the fundamentals of Data Science and the challenges of complex data sets (for example big data).” The interview date is November 27, 2014. All details available here.
Archive for England
Great poster session yesterday night and at lunch today. Saw an ABC poster (by Dennis Prangle, following our random forest paper) and several MCMC posters (by Marco Banterle, who actually won one of the speed-meeting mini-project awards!, Michael Betancourt, Anne-Marie Lyne, Murray Pollock), and then a rather different poster on Mondrian forests, that generalise random forests to sequential data (by Balaji Lakshminarayanan). The talks all had interesting aspects or glimpses about big data and some of the unnecessary hype about it (them?!), along with exposing the nefarious views of Amazon to become the Earth only seller!, but I particularly enjoyed the astronomy afternoon and even more particularly Steve Roberts sweep through astronomy machine-learning. Steve characterised variational Bayes as picking your choice of sufficient statistics, which made me wonder why there were no stronger connections between variational Bayes and ABC. He also quoted the book The Fourth Paradigm: Data-Intensive Scientific Discovery by Tony Hey as putting forward interesting notions. (A book review for the next vacations?!) And also mentioned zooniverse, a citizens science website I was not aware of. With a Bayesian analysis of the learning curve of those annotating citizens (in the case of supernovae classification). Big deal, indeed!!!
I had quite a special day today as I travelled through Birmingham, made a twenty minutes stop in Coventry to drop my bag in my office, went down to London to collect a most kindly loaned city-bike and took the train back to Coventry with the said bike… On my way from Bristol to Warwick, I decided to spend the night in downtown Birmingham as it was both easier and cheaper than to find accommodation on Warwick campus. However, while the studio I rented was well-designed and brand-new, my next door neighbours were not so well-designed in that I could hear them and the TV through the wall, despite top-quality ear-plugs! After a request of mine, they took the TV off but kept to the same decibel level for their uninteresting exchanges. In the morning I tried to go running in the centre of Birmingham but, as I could not find the canals, I quickly got bored and gave up. As Mark had proposed to lend me a city bike for my commuting in [and not to] Warwick, I then decided to take the opportunity of a free Sunday to travel down to London to pick the bike, change the pedals in a nearby shop, add an anti-theft device, and head back to Coventry. Which gave me the opportunity to bike in London by Abbey Road, Regent Park, and Hampstead, before [easily] boarding a fast train back to Coventry and biking up to the University of Warwick campus. (Sadly to discover that all convenience stores had closed by then… )
Last and maybe most exciting day of the “High-dimensional Stochastic Simulation and Optimisation in Image Processing” in Bristol as it was exclusively about simulation (MCMC) methods. Except my own talk on ABC. And Peter Green’s on consistency of Bayesian inference in non-regular models. The talks today were indeed about using convex optimisation devices to speed up MCMC algorithms with tools that were entirely new to me, like the Moreau transform discussed by Marcelo Pereyra. Or using auxiliary variables à la RJMCMC to bypass expensive Choleski decompositions. Or optimisation steps from one dual space to the original space for the same reason. Or using pseudo-gradients on partly differentiable functions in the talk by Sylvain Lecorff on a paper commented earlier in the ‘Og. I particularly liked the notion of Moreau regularisation that leads to more efficient Langevin algorithms when the target is not regular enough. Actually, the discretised diffusion itself may be geometrically ergodic without the corrective step of the Metropolis-Hastings acceptance. This obviously begs the question of an extension to Hamiltonian Monte Carlo. And to multimodal targets, possibly requiring as many normalisation factors as there are modes. So, in fine, a highly informative workshop, with the perfect size and the perfect crowd (which happened to be predominantly French, albeit from a community I did not have the opportunity to practice previously). Massive kudos to Marcello for putting this workshop together, esp. on a week where family major happy events should have kept him at home!
As the workshop ended up in mid-afternoon, I had plenty of time for a long run with Florence Forbes down to the Avon river and back up among the deers of Ashton Court, avoiding most of the rain, all of the mountain bikes on a bike trail that sounded like trail running practice, and building enough of an appetite for the South Indian cooking of the nearby Thali Café. Brilliant!
After a nice morning run down Leigh Woods and on the muddy banks of the Avon river, I attended a morning session on hyperspectral image non-linear modelling. Topic about which I knew nothing beforehand. Hyperspectral images are 3-D images made of several wavelengths to improve their classification as a mixture of several elements. The non-linearity is due to the multiple reflections from the ground as well as imperfections in the data collection. I found this new setting of clear interest, from using mixtures to exploring Gaussian processes and Hamiltonian Monte Carlo techniques on constrained spaces… Not to mention the “debate” about using Bayesian inference versus optimisation. It was overall a day of discovery as I am unaware of the image processing community (being the outlier in this workshop!) and of their techniques. The problems mostly qualify as partly linear high-dimension inverse problems, with rather standard if sometimes hybrid MCMC solutions. (The day ended even more nicely with another long run in the fields of Ashton Court and a conference diner by the river…)