Archive for book review

A Closed and Common Orbit

Posted in Statistics with tags , , , , , , , on February 27, 2018 by xi'an

This book by Becky Chambers comes as a sequel of sorts to her first [science-fiction] book, A Long Way to a Small Angry Planet. Book that I liked a lot for its construction of relationships between the highly different team members of a spaceship. In this new book, the author pursues a similar elaboration of unlikely friendships between human and alien species, and AIs. If the first book felt homey, this one is even more so, with essentially two principal characters followed alternatively throughout the book, until the stories predictably cross. It is fairly well-written, with again a beautiful cover, but I cannot say it is as magisterial as the first book. The book-long considerations on the nature of AI and of cloned humans are certainly interesting and deep enough, but the story tension ebbs at time, especially for the story in the past since we know from the beginning that the main character will reappear in the current time. Not reaching the superlatives of a Hugo or Clarke Award in my opinion (albeit nominated for these prizes). Still a most enjoyable read!

la belle sauvage [book review]

Posted in Statistics with tags , , , , , , , , , , , , , , , on February 25, 2018 by xi'an

Another book I brought back from Austin. And another deeply enjoyable one, although not the end of a trilogy of trilogies this time. This book, La Belle Sauvage, is first in a new trilogy by Philip Pullman that goes back to the early infancy of the hero of His Dark Materials, Lyra. Later volumes will take place after the first trilogy.

This is very much a novel about Oxford, to the point it sometimes seems written only for people with an Oxonian connection. After all, the author is living in Oxford… (Having the boat of the two characters passing by the [unnamed] department of Statistics at St. Giles carried away by the flood was a special sentence for me!)

Also, in continuation of His Dark Materials, a great steampunk universe, with a very oppressive Church and so far a limited used of magicks! Limited to the daemons, again in continuation with past volumes…

Now, some passages of the book remind me of Ishiguro’s buried giant, in the sense that the characters meeting myths from other stories may “really” meet them or instead dream. This is for instance the case when they accost at a property where an outworldy party is taking place and no-one is noticing them. Or when they meet a true giant that is a river deity, albeit not in the spirit of Ben Aaronovitch’s Rivers of London novels.

The story is written in the time honoured setup of teenager discovery travels, with not so much to discover as the whole country is covered by water. And the travel gets a wee bit boring after a while, with a wee bit too many coincidences, the inexplicable death (?) of a villain, and an hurried finale, where the reverse trip of the main characters takes a page rather than one book…

Trivia: La Belle Sauvage was also the name of the pub in Ludgate Hill where Pocahontas and her brother Tomocomo stayed when they first arrived in London. And The Trout is a true local pub, on the other side of Port Meadow [although I never managed to run that far in that direction while staying in St. Hugh, Oxford, last time, the meadow being flooded!].

Looking forward the second volume (already written, so no risk of The Name of the Wind or Game of Thrones quagmires, i.e., an endless wait for the next volume!), hoping the author keeps up the good work, the right tension in the story, and avoids by all means parallel universes, which were so annoying in the first trilogy! (I do remember loosing interest in the story during the second book and having trouble finishing the third one. I am not sure my son [who started before me] ever completed the trilogy…)

Assassin’s fate [book review]

Posted in Books, Kids with tags , , , , , on February 3, 2018 by xi'an

I am afraid it is impossible to report on Assassin’s Fate without introducing spoilers, so would-be readers, be warned! The end of a long if enjoyable journey with FitzChivalry, the royal Assassin created by Robin Hobb two decades ago, in a book that brings together almost all characters she introduced in the five trilogies linking the Six Duchies, the Rain Wild, and beyond. Beyond the imperfections of some slow-pace sections and of the infuriating stubbornness of Fitz along with his righteousness at times, the conclusion of the book is stunning and perfectly closes the series, leaving the reader who has followed these characters for years and enjoyed the carefully constructed universe behind them, as well as the psychological depth of most of them, with a peaceful and bittersweet sadness. Never have so many owed so much to so few, some would add about Fitz, The Fool and Bee, given the upheaval they bring to this whole universe, impacting first and foremost the Liveship traders… In my opinion, the saga of FitzChivalry that concludes with this book stands among the most realised and elaborated ones in fantasy, primarily for its highly attaching (and far from heroic) characters. Who definitely belong to a pantheon of fantasy characters that one remembers along the years, even with long interruptions.

humanism [book review]

Posted in Books, Kids with tags , , , , , on January 14, 2018 by xi'an

Along Atheism a very short introduction, I also bought Humanism a very short introduction, as they come by two at the Warwick campus bookstore (!). And here is a very short review.

Written by Stephen Lee, the book is much less irritating than Atheism. In my opinion. Maybe because it is constructed in a much more positive way, maybe because the quotes and illustrations suited me better, maybe because it was another day, or maybe because the stress on the “human” rather than on the “a-” is closer to my own philosophy. Still, the core of the two books is essentially the same, namely a rebuke of the argument that morality only comes as a byproduct of religion(s), and a rather standard processing of arguments for and against the existence of god(s). Plus entries on humanist education and the meaning of life. And a nice cover. Pleasant but not earth-breaking to the point of convincing sceptics.

the shockwave rider [book review]

Posted in Statistics with tags , , , , , , on January 11, 2018 by xi'an

I ordered this book from John Brunner when I found this was the precursor to Neuromancer and the subsequent cyberpunk literature. And after reading it during the Xmas break I am surprised it is not more well-known. Indeed, the plot, the style, the dystopian society in The Shockwave Rider all are highly original, and more “intellectual” than successors like Neuromancer or Snow Crash. Reading this 1975 book forty years later also reveals its premonitory features, from inventing the concept of computer worm (along with a pretty accurate description), to forecasting (or being aware of plans for) cell-phones, the Net, the move to electric cars, and Wikipedia, with the consequence of being always visible for whoever controls the network. The characters are flawed in that they are too charicaturesque, but this is somewhat secondary since the main appeal of the book is to discuss the features of an all-connected world. And the way to recover power to the people against a government controlling the network and the associated data. The time being the 1970’s the resolution via a hippie commune in Northern California (like Eureka!) is a bit outdated and definitely “rosy”, and does not foresee the issue of “digital democracy” being threatened by a strong polarisation into estranged communities, but I still enjoyed the book tremendously. (As a bonus, I got the first edition of the book at a ridiculous price! With this somewhat outdated cover.)

ready player one [book review]

Posted in Books, Kids, pictures with tags , , , , , , , , on January 2, 2018 by xi'an

This book was presumably suggested to me by an Amazon AI based on my previous browsing, and I got intrigued enough by the summary plot and the above cover to order it while in Austin and read it on the way back to Paris. The setting of the story is a catastrolyptic near-future (2044) where gas is a luxury and most of the planet is unemployed and spends its time in a all-immersive free-access virtual reality. Five teenagers join million others in a quest to win a fortune… Against an evil corporation that seeks this fortune and control of the virtual universe. The story revolves around this quest, with some forays in the real world, and a reflection on characters who only know each other via their avatars. Other books based on videogaming come to mind, from Ender’s Game to Neuromancer, to Diamond Age, to REAMDE… But this one is much more focussed on the nature of video-gaming and on the feature of such of a society. I enjoyed the book to the point of staying up late for several evenings in a row, even though the plot is somewhat weak at the societal level, i.e. in describing the economic dynamics of such a society, but setting the games, movies and music themes within the 80’s is obviously catering to readers like me (although I miss a large part of the references). The book was published in 2011, so this is not any recent publication, but there is a movie by Steven Spielberg coming out soon.

about paradoxes

Posted in Books, Kids, Statistics, University life with tags , , , , , , , , , , on December 5, 2017 by xi'an

An email I received earlier today about statistical paradoxes:

I am a PhD student in biostatistics, and an avid reader of your work. I recently came across this blog post, where you review a text on statistical paradoxes, and I was struck by this section:

“For instance, the author considers the MLE being biased to be a paradox (p.117), while omitting the much more substantial “paradox” of the non-existence of unbiased estimators of most parameters—which simply means unbiasedness is irrelevant. Or the other even more puzzling “paradox” that the secondary MLE derived from the likelihood associated with the distribution of a primary MLE may differ from the primary. (My favourite!)”

I found this section provocative, but I am unclear on the nature of these “paradoxes”. I reviewed my stat inference notes and came across the classic example that there is no unbiased estimator for 1/p w.r.t. a binomial distribution, but I believe you are getting at a much more general result. If it’s not too much trouble, I would sincerely appreciate it if you could point me in the direction of a reference or provide a bit more detail for these two “paradoxes”.

The text is Chang’s Paradoxes in Scientific Inference, which I indeed reviewed negatively. To answer about the bias “paradox”, it is indeed a neglected fact that, while the average of any transform of a sample obviously is an unbiased estimator of its mean (!), the converse does not hold, namely, an arbitrary transform of the model parameter θ is not necessarily enjoying an unbiased estimator. In Lehmann and Casella, Chapter 2, Section 4, this issue is (just slightly) discussed. But essentially, transforms that lead to unbiased estimators are mostly the polynomial transforms of the mean parameters… (This also somewhat connects to a recent X validated question as to why MLEs are not always unbiased. Although the simplest explanation is that the transform of the MLE is the MLE of the transform!) In exponential families, I would deem the range of transforms with unbiased estimators closely related to the collection of functions that allow for inverse Laplace transforms, although I cannot quote a specific result on this hunch.

The other “paradox” is that, if h(X) is the MLE of the model parameter θ for the observable X, the distribution of h(X) has a density different from the density of X and, hence, its maximisation in the parameter θ may differ. An example (my favourite!) is the MLE of ||a||² based on x N(a,I) which is ||x||², a poor estimate, and which (strongly) differs from the MLE of ||a||² based on ||x||², which is close to (1-p/||x||²)²||x||² and (nearly) admissible [as discussed in the Bayesian Choice].