a journal of the plague year [lazy August reviews]

Read Blood of Empire, the final volume in the Gods of Blood and Powder trilogy. By Brian McClellan. Which I enjoyed reasonably well as bedside literature, although its weight meant it would fall at the slightest hint of sleep… It took me longer than expected to connect to the story, given I had read the previous volume a few months ago. This series is classified as “flintrock fantasy”, a category I had never heard of previous, meaning a limited amount of gunpower is used in weapons, along with the aid of magical abilities (for the happy few). The style is a wee bit heavy and repetitive, but the characters are definitely engaging if over-prone to inner dialogues… The only annoying part in the plot is the presence of a super-evil character about to be become a god, which ruins most of the balance in the story.

Had a long-pending due watch at Trainspotting T2. (Loved the NYT label as “Rated R for a bagful of vomit, mouthfuls of bigotry and nosefuls of cocaine”, obviously in the same regressive spirit as the film.) This is definitely a sequel to the first film. And hence hardly comprehensible on its own. Except for a few locations like a run-down pub on the edge of nowhere, a flat overlooking a car part dump and Spud’s high-rise welfare housing, T2 lacks the gritty vision of Edinburgh found in its forbear. And the characters have lost their toxic edge, except maybe very much maybe for the psychopath Franck. Even the de rigueur final swindle has a rosy and predictable justification. Fun nonetheless! On the (tourist) side, I enjoyed a mostly superfluous scene where Renton takes Spud running up Arthur’s Seat along its most scenic route, with an iconic end image of Edinburgh gradually fading into fog. There is also a surreal (short) scene on Rannoch Mor, with the Oban train stopping at the hikers’ stop. (I never managed to start Welsh’s books, due to their phonetic rendering of Edniburghian Scots that make reading unbearable..! By comparison, most dialogues are understandable. A funny line when the hostess welcoming tourists at Edinburgh Airport with a mock Scottish accent acknowledges she is from Slovakia.) Camera tricks like fast backward and colour filters a wee bit old-fashioned and heavy-handed, in the spirit of the first movie as if nothing had ever happened since. Maybe the moral of the story. Not looking for a potential T3, though.

Read a forgotten volume in the Bernhard Günther series of Philip Kerr, A man without breath. As usual building on historical events from Nazi Germany to set this ambivalent character at the centre of the action, which is this time the discovery and exploitation of the Katyǹ massacres by the Nazi propaganda to drive an edge between the Soviet Union and the other Allies. The book is rather uneven, with too many plots, subplots, and characters, and open criticisms of the Nazi regime between complete strangers do not ring particularly realistic. And draw attention away from their own massacres, like Babi Yar (celebrated in Dmitri Shostakovitch’s Symphony No. 13). Interestingly, given that I read the book at the time of the JSM round-table, a thread in the story links to the Spanish Civil War and the attempt by fascist doctors like Vallejo Nágera to picture left-wing Spaniards as psychiatrically degenerates, fantasying the existence of a “red” gene… (It took me a while to trace the reference in the title to Goebbels’ quote “A nation with no religion is like a man without breath.” )

at the centre of Bayes

ABC variable selection

Prior to the ISBA 2018 meeting, Yi Liu, Veronika Ročková, and Yuexi Wang arXived a paper on relying ABC for finding relevant variables, which is a very original approach in that ABC is not as much the object as it is a tool. And which Veronika considered during her Susie Bayarri lecture at ISBA 2018. In other words, it is not about selecting summary variables for running ABC but quite the opposite, selecting variables in a non-linear model through an ABC step. I was going to separate the two selections into algorithmic and statistical selections, but it is more like projections in the observation and covariate spaces. With ABC still providing an appealing approach to approximate the marginal likelihood. Now, one may wonder at the relevance of ABC for variable selection, aka model choice, given our warning call of a few years ago. But the current paper does not require low-dimension summary statistics, hence avoids the difficulty with the “other” Bayes factor.

In the paper, the authors consider a spike-and… forest prior!, where the Bayesian CART selection of active covariates proceeds through a regression tree, selected covariates appearing in the tree and others not appearing. With a sparsity prior on the tree partitions and this new ABC approach to select the subset of active covariates. A specific feature is in splitting the data, one part to learn about the regression function, simulating from this function and comparing with the remainder of the data. The paper further establishes that ABC Bayesian Forests are consistent for variable selection.

“…we observe a curious empirical connection between π(θ|x,ε), obtained with ABC Bayesian Forests  and rescaled variable importances obtained with Random Forests.”

The difference with our ABC-RF model choice paper is that we select summary statistics [for classification] rather than covariates. For instance, in the current paper, simulation of pseudo-data will depend on the selected subset of covariates, meaning simulating a model index, and then generating the pseudo-data, acceptance being a function of the L² distance between data and pseudo-data. And then relying on all ABC simulations to find which variables are in more often than not to derive the median probability model of Barbieri and Berger (2004). Which does not work very well if implemented naïvely. Because of the immense size of the model space, it is quite hard to find pseudo-data close to actual data, resulting in either very high tolerance or very low acceptance. The authors get over this difficulty by a neat device that reminds me of fractional or intrinsic (pseudo-)Bayes factors in that the dataset is split into two parts, one that learns about the posterior given the model index and another one that simulates from this posterior to compare with the left-over data. Bringing simulations closer to the data. I do not remember seeing this trick before in ABC settings, but it is very neat, assuming the small data posterior can be simulated (which may be a fundamental reason for the trick to remain unused!). Note that the split varies at each iteration, which means there is no impact of ordering the observations.

go, Iron scots!

ABC in Ed’burgh

A glorious day for this new edition of the “ABC in…” workshops, in the capital City of Edinburgh! I enjoyed very much this ABC day for demonstrating ABC is still alive and kicking!, i.e., enjoying plenty of new developments and reinterpretations. With more talks and posters on the way during the main ISBA 2018 meeting. (All nine talks are available on the webpage of the conference.)

After Michael Gutmann’s tutorial on ABC, Gael Martin (Monash) presented her recent work with David Frazier, Ole Maneesoonthorn, and Brendan McCabe on ABC  for prediction. Maybe unsurprisingly, Bayesian consistency for the given summary statistics is a sufficient condition for concentration of the ABC predictor, but ABC seems to do better for the prediction problem than for parameter estimation, not losing to exact Bayesian inference, possibly because in essence the summary statistics there need not be of a large dimension to being consistent. The following talk by Guillaume Kon Kam King was also about prediction, for the specific problem of gas offer, with a latent Wright-Fisher point process in the model. He used a population ABC solution to handle this model.

Alexander Buchholz (CREST) introduced an ABC approach with quasi-Monte Carlo steps that helps in reducing the variability and hence improves the approximation in ABC. He also looked at a Negative Geometric variant of regular ABC by running a random number of proposals until reaching a given number of acceptances, which while being more costly produces more stability.

Other talks by Trevelyan McKinley, Marko Järvenpää, Matt Moores (Warwick), and Chris Drovandi (QUT) illustrated the urge of substitute models as a first step, and not solely via Gaussian processes. With for instance the new notion of a loss function to evaluate this approximation. Chris made a case in favour of synthetic vs ABC approaches, due to degradation of the performances of nonparametric density estimation with the dimension. But I remain a doubting Thomas [Bayes] on that point as high dimensions in the data or the summary statistics are not necessarily the issue, as also processed in the paper on ABC-CDE discussed on a recent post. While synthetic likelihood requires estimating a mean function and a covariance function of the parameter of the dimension of the summary statistic. Even though estimated by simulation.

Another neat feature of the day was a special session on cosmostatistics with talks by Emille Ishida and Jessica Cisewski, from explaining how ABC was starting to make an impact on cosmo- and astro-statistics, to the special example of the stellar initial mass distribution in clusters.

Call is now open for the next “ABC in”! Note that, while these workshops have been often formally sponsored by ISBA and its BayesComp section, they are not managed by a society or a board of administrators, and hence are not much contrived by a specific format. It would just be nice to keep the low fees as part of the tradition.