Бессонные Хроники
[Most Recent Entries]
[Calendar View]
[Friends]
Below are the 20 most recent journal entries recorded in
Kurumo's LiveJournal:
[ << Previous 20 ]
| Tuesday, April 19th, 2011 | | 12:42 am |
| | Tuesday, September 21st, 2010 | | 10:39 pm |
The following problem occurred to me on the way home: Suppose you are a navigator on a ship. You take three chronometers with you (call them A, B and C), setting them to the same time at the beginning of a voyage. Thereafter you make daily observations, and after some number of days (greater than 30) estimate that in one full day chronometer B tends to run three seconds fast on average, with a standard deviation of 1 second. Let's assume that the other two chronometers show the same time to within your measurements' precision, and that the fast chronometer is reset daily to the time measured via the other two. Thereafter you are injured, and cannot make measurements for N days. Upon your recovery you observe that all three chronometers display different times (call them Ta, Tb, Tc). What is the best estimate of the true time? What is the uncertainty in the estimate? Haven't solved it yet, but seems like this may be nontrivial. If someone had seen a problem like this before, please let me know. | | Thursday, May 20th, 2010 | | 4:51 pm |
| | Tuesday, March 9th, 2010 | | 1:48 am |
Welcome to the future department writes... that if I read this correctly, these people have very convincingly controlled (I hesitate to say 'cured') HIV and leukemia, at the same time,
via a stem cell transplant from a CCR5-deleted donor: http://content.nejm.org/cgi/content/full/360/7/692
I quote:
"The second procedure led to a complete remission of the acute myeloid leukemia, which was still in remission at month 20 of follow-up."
"... in this patient, no active, replicating HIV could be detected 20 months after HAART had been discontinued."
That's some impressive work.
This was a year ago, incidentally. | | Tuesday, March 2nd, 2010 | | 6:25 pm |
Tales from interviews Me: It helps to draw this picture. Consider a rectangle with a rectangular hole in it. The hole may be placed anywhere inside the rectangle in arbitrary orientation. The only tool you have is a ruler without markings, that is, you cannot measure distances or angles. We would like to draw a (single) line that divides this figure in two parts of equal area.
Candidate (MSc in Mechanical Engineering): <spends fifteen minutes thinking and requests a hint>
Me: Suppose we just want to divide a single rectangle in two parts of equal area. How can we do that with a straightedge?
Candidate: Draw a diagonal.
Me: How else?
Candidate: Draw the other diagonal.
Me: And?
Candidate: Draw a line parallel to the sides through the center.
Me (losing patience): ... and the other such line, yes. What other ways are there to divide a rectangle in two equal parts?
Candidate: There are no other ways.
Second interviewer: <restrains himself physically from swearing>
Me: Wait, let me get this straight. So if we draw an arbitrary line though the center, it will not divide a rectangle in equal parts.
Candidate: No.
Me: <smashes head againt the wall repeatedy>
Second interviewer: I need a drink. Hilarity ensues.
The end. P.S. I may have told this story before; the reason I put it up is because this happened the second time. | | Sunday, September 27th, 2009 | | 2:55 pm |
Once again from the "Welcome to the future..." department. Segway is all well and good, but this is pure future shock: Luke armThey have a neural interface at this point. The arm is in wide clinical trials with the US military. | | Thursday, September 10th, 2009 | | 8:51 pm |
From the 'Welcome to the future' department... Using Neural Measures of Economic Value to Solve the Public Goods Free-Rider Problem. http://www.sciencemag.org/cgi/content/abstract/1177302Briefly, these people apply pattern classification algorithms to fMRI data to determine whether the value a person assigns to something is what he or she actually believes the value to be. Not quite a lie detector, but conceptually in the same category. Anybody else find this development mildly disturbing? | | Monday, August 31st, 2009 | | 11:54 am |
Лилипуты... http://news.bbc.co.uk/2/hi/technology/8226509.stm"Alan Turing is most famous for his code-breaking work at Bletchley Park during WWII, ... However he also made significant contributions to the emerging fields of artificial intelligence and computing." "However"? | | Thursday, August 13th, 2009 | | 12:37 am |
"Six is a number perfect in itself, and not because God created the world in six days; rather the contrary is true. God created the world in six days because this number is perfect, and it would remain perfect, even if the work of the six days did not exist." St. Augustine, 'De Civitate Dei' | | Sunday, June 28th, 2009 | | 5:43 pm |
| | Tuesday, March 10th, 2009 | | 1:00 am |
| | Saturday, February 28th, 2009 | | 11:18 pm |
Suppose you are given a language called C--. It is just like C, except there is no if, no switch, and no tertiary operator: i.e. no conditionals of any kind. Can we write all the same programs in it as in, say, C or C++, and if so, how? To make the problem more interesting, suppose we also remove all the loop constructs? I think I will ask this in interviews. | | Wednesday, February 25th, 2009 | | 12:39 am |
Tales from interviews Me: Please describe an algorithm you would use to search for a particular value in a sorted array. Candidate (MS in CS): <in fits and starts, describes and writes pseudo code for binary search> Me: What is the running time of this algorithm in terms of O-notation? Candidate: <brightly> n^2 Me: <falling of a chair> Pardon?.. Candidate: no, wait.. That would have to be 2^n Me: <chokes on coffee> Fini | | Friday, February 13th, 2009 | | 12:59 am |
A puzzle with a most elegant solution. Two people play a game. On your turn, you pick up a number 1 through 9 without replacement. To win, one needs to collect three numbers that sum to 15. Describe the optimal strategy. | | Wednesday, February 11th, 2009 | | 1:38 am |
Tales from interviews Me: Please describe a class design for a deck of cards, say for something like an online casino. Candidate: <very quickly produces a design in which Players publicly inherit from Cards> Me: ... ?!?.. Немая сцена | | Sunday, February 8th, 2009 | | 11:41 pm |
Came upon this. Lots of short interviews with some of the people who made the 20th century. Hans Bethe, Mandelbrot, Don Knuth, Gell-Mann, John Wheeler, Josh Maynard Smith, Francis Crick, Edward Teller, Atiyah, etc. http://www.peoplesarchive.com/home.jsp | | Wednesday, December 17th, 2008 | | 2:47 pm |
Observations on interviewing: - over 3/4 of all candidates with C/C++ on the resume cannot reverse a linked list in those languages (in place)
- over 1/3 cannot do it in principle, in any language or by any method, with multiple tries
- over 1/2 have no concept of how to reverse digits in a number without converting the said number to a string
- the above holds regardless of the degree; if anything, Masters in Computer Science do worse on average. (I regard MS as a red flag at this point).
- candidates who declare Java as their preferred programming language in general have a weaker grasp of algorithms and little to no idea of architecture :( | | Monday, December 15th, 2008 | | 1:15 pm |
In Febraury of last year, when I was looking for a job, I passed my resume to somebody at Akamai at a career fair. Yesterday I got a letter from them: "We got your resume and would like to invest time (sic.) in evaluation of your skills." I think it might be some sort of record. A year and ten months. | | Sunday, December 7th, 2008 | | 5:03 pm |
Who'd have thought it possible. "In recent work that is nothing short of spectacular, X. Leroy has developed a formally verified compiler for the C programming language." Thomas C. Hales, "Formal Proof", Notices of the AMS, December 2008 (55, 11), p. 1377 | | Monday, November 24th, 2008 | | 12:31 am |
A source of morality, indeed. Against religion( I quote: )Granted, correlation is not causation, and the statistical analysis in the article is nonexistent. But even the correlations are interesting, and are a touch counterintuitive to people. Also, Jensen, Journal of Religion & Society 8: 1–14. |
[ << Previous 20 ]
|
![[info]](http://l-stat.livejournal.com/img/userinfo.gif?v=92.1){kind=small}
angrauko's journal