Archive for May 2009
… is the title of a great article in the Financial Times from 04/29/09 (thx for pointing me to that one). It is about the origins of the copula formula and its connection to actuarial science where a similar problem, to estimate or quantify the remaining life expectancy in couples, occurs – more specifically the remaining life expectancy of the left-behind partner conditional on the other one being dead. The broken heart syndrom:
Pages and pages of death records showed the same marked trend: that in human couples, the death of one partner significantly increases the chances of the death of the other. Dying of a broken heart – in the most general sense, not necessarily from stress cardiomyopathy – was not a rare occurrence, but something of a statistical probability.
A nice story about 1001 correlations (covering everything from Black-Scholes-Merton, to LTCM, to Russian Bonds, etc) … Definitely worth reading.
A wired article from the 4/29/09 titled ‘Google Could Have Caught Swine Flu Early‘ suggests that google’s search data could have been used as an early warning system in order to detect flu activity. Although a good idea the problem so far seems to be that it is not clear which combinations or search patterns to monitor:
But the Google Flu Trends team, which aggregates and analyzes search queries to estimate how many people are sick, wasn’t watching Mexican flu data until after the outbreak had already begun. That highlights the problem with tech-heavy disease-detection systems: Often, we don’t know what internet data to look at until after a problem starts.
The early signals of disease are hidden in plain sight, and it takes humans recognizing that something is happening before the computers can be asked to find it. And even if Flu Trends had picked up a noticeable bump in flu searches in Mexico early, a lot of additional analysis would have been required to understand the potential severity of the pandemic.
While showing that a lot of data can contain a lot of information, at the same time it unfortunately also highlights the problem that more data is not necessarily more information if you do not know what to look for. Anyways, enjoy the article!
Update (05/04/2009): There is also a nice visualization on ushahidi available that crowd-sources cases and highlights their locations.
Another article worthwhile reading: “Don’t Blame The Elite” by Tim Harford.
For years, we were told that Wall Street attracted the very best. That was why American investment banks were the envy of the world; that was why stratospheric salaries and bonuses were essential. Other financial centers, such as London, fought tooth and nail to attract the same elite. They were worth it, we were told: If you pay peanuts, you get monkeys.
By accident I stumbled upon the following ad/video for the NextGen gate to gate system that is supposed to make air travel more efficient. Basically, integrated information technology should streamline communication and reduce overhead that arose from decentralized management. As a result planes should e.g., fly closer together “without any loss of safety”… But have a look yourself. If you are interested you will find more details on the FAA NextGen pages.
A new version of the GNU Linear Programming Kit has been released. The version 4.38 includes a few improvements in the dual simplex routines and two new packages were included into the interior-point solver for reordering the matrices prior to the Cholesky factorization (for more information see SuiteSparse). Expect a new version of GUSEK soon including the new glpsol executable.
Release date: May 02, 2009
GLPK (GNU Linear Programming Kit) is intended for solving large-scale
linear programming (LP), mixed integer linear programming (MIP), and
other related problems. It is a set of routines written in ANSI C and
organized as a callable library.
In this release:
API routines glp_read_mps and glp_write_mps were improved.
Some improvements were made in the dual simplex routines.
Two external software modules AMD and COLAMD were included in
the distribution (for more details please see src/amd/README
and src/colamd/README). Now they are used in the interior-point
solver to reorder the matrix prior to Cholesky factorization.
API routine glp_ipt_status may return two new statuses due to
changes in the routine glp_interior. For details please see the
reference manual included in the distribution.
A minor bug was fixed in the graph/network routines. Thanks to
Nelson H. F. Beebe <email@example.com> for bug report.
See GLPK web page at <http://www.gnu.org/software/glpk/glpk.html>.
GLPK distribution can be ftp’ed from <ftp://ftp.gnu.org/gnu/glpk/> or
from some mirror ftp sites; see <http://www.gnu.org/order/ftp.html>.