Books to Read While the Algae Grow in Your Fur, September 2008
Creations: Aberrant Ideas of Human Origins from Ancient Astronauts to Aquatic
- Donna Kossy is one of the best writers about weird beliefs around. (Go
right now if you haven't.) This is a look some ideas about where humanity came
from, and where it might go, which are currently on the fringes of
respectability at best, although in some cases (creationism, eugenics) they
didn't use to be. She's particularly good, here, on the influence of
Theosophy. As always, Kossy tries very hard to be sympathetic to those she
reports on, even in cases where that's obviously very hard (like eugenicists).
Her readers, however, are under no such obligations.
- (She misses
astronaut connection, however; in a sense the Cthulhu Mythos is what you
get when you take Blavatsky's elaborate scheme of pre-human,
material-life-but-not-as-we-know-it races, and add the assumption that aliens
will be alien.)
- Kage Baker, The House of the
- Another intricately-plotted fantasy
novel, including a lengthy section where Epic Fantasy is depicted as a kind
Also, a portrait of the saint as a young woman.
- — Brad DeLong
a long excerpt of one of the more amusing scenes.
- — Later book in the same world
(only loosely connected).
- Warren Ellis, Chris Weston
- Alternate history of the British conquest of space, aided by
Wernher "we aim for the
stars, and sometimes we hit
Braun. The technology thus closesly follows von Braun's actual
Johnson and Warren
A Graphic Mystery
- Historical mystery about lynching and passing.
- Denis Bosq
Blanke, Inference and Prediction in Large Dimensions
- Some aspects of prediction problems in the presence of either
infinite-dimensional data (such as continuous processes observed over finite
intervals of time), or infinite-dimensional parameters (such as arbitrary
smooth probability distributions for noise), or both. The general setting is
that of mixing processes; the major tools are (1) limit theorems for mixing
processes, especially exponential inequalities; (2) kernel estimators (as in
Bosq's earlier book
Statistics for Stochastic Processes); and (3) projection estimators,
i.e., projecting the data, suitably massaged, on to a countable orthonormal
basis in an appropriate Hilbert space, and adaptively truncating small
components. The authors refer to all such basis decompositions as "Fourier
analysis", whether the basis functions are trigonometric functions or not.
Much of this comes together in the last few chapters, on "functional linear
processes", i.e., linear processes taking values in infinite-dimensional
function spaces, such as those of continuous curves. As is often the case, by
moving to Hilbert space one can linearize an otherwise non-linear
time-evolution, and so do a lot more than you'd think with autoregressive
- This is very much a book for advanced students; familiarity with
treatment of stochastic processes is essential, as is a good grasp of
ordinary, finite-dimensional estimation theory, and of course familiarity with
Fourier analysis and Hilbert space methods. (Knowledge of ordinary time-series
methods is not so much a requirement.) The exposition is very dense and
sometimes confusing, though only one section is actually impenetrable. (This
is the tangential one on Blackwell's method for forecasting the probabilities
of arbitrary events; it's a hard topic, but the discussion
in Prediction, Learning and
Games really is infinitely clearer.) Many of the results are either
new to this book, or might as well be new.
Lynch, The Lies of Locke Lamora
Seas Under Red Skies
- Mind-candy: a weird yet very entertaining hybrid of fantasy and crime-caper
novel. — Sequel.
Books to Read While the Algae Grow in Your Fur;
Scientifiction and Fantastica;
Pleasures of Detection, Portraits of Crime;
Enigmas of Chance;
Posted at September 30, 2008 23:59 | permanent link