Forgot your password?

typodupeerror

Comment: Most published research findings are false (Score 4, Informative) 233

by kahizonaki (#39596909) Attached to: Majority of Landmark Cancer Studies Cannot Be Replicated
A recent PLOS article (free to view!) analyses modern research, coming to the conclusion that most research findings are false.
TLDR: Because of the nature of the statistics used and the fact that only positive results are reported.
http://www.plosmedicine.org/article/info:doi/10.1371/journal.pmed.0020124

Comment: Re:Buses US only? (Score 2) 187

by kahizonaki (#39121617) Attached to: How Google Is Remapping Public Transportation
Err, I posed as AC a second ago, and forgot to post the link :) Anyways, all Japanese people I know use this site to route (mostly between train stations?), but it gives you all things including normal buses, high speed buses, shinkansen, walking, water ferries, etc. I don't know if they have an english version though... http://transit.loco.yahoo.co.jp/

+ - Kim Jun-Il dead->

Submitted by kahizonaki
kahizonaki writes "In Japan right now all the television stations are reporting the death of Kim Jung-Il (the de facto dictator of North Korea), which was apparently announced officially within North Korea earlier as well."
Link to Original Source

Comment: Ridiculous (Score 1) 760

by kahizonaki (#34597750) Attached to: 'YouCut' Targets National Science Foundation Budget
Disclaimer: I am funded by the NSF.

First of all, having people who probably have little-to-no scientific training, let alone any training or expertise in the field that the grant is in, decide whether a particular research project is "a waste of time" is beyond stupidity. It is equivalent to allowing the average american to micro-manage troop movements on the war-zone, allocate ad hoc rations/supplies to each region of the world, etc. In other words, the solutions provided by the "people" will be far from efficient, far from optimal, and indeed probably just dead *wrong* (i.e. soldiers would starve because people thought "nah they don't need this cooking fuel there, they can use firewood!" or something along equally stupid lines.

Another example would be letting people decide civil engineering matters. Like, let's use the cheap steel for the bridge, it's good enough! Or, let's route this highway right through over there, look it's wide open! -- without understanding all the effects and repercussions that taking any of these actions would have (which a properly trained civil engineer etc. would be more likely to recognize).

Of course, one of the explicit stated purposes of the NSF is to broaden appreciation and understanding of science. It's so important that it's almost impossible to get a grant without being able to convincingly show that your project will have broader impacts outside of your subfield. Of course, the people who would be going through the grants by this YouCut thing wouldn't understand why certain seemingly retarded research projects are important (e.g. why bother measuring the weight of the Earth's core, who cares? I want a new car and a TV.) when really it could be a very serious question that many other projects hinge upon (e.g. geothermal energy, satellites that might be affected by the magnetic field that is generated by the core, etc...). Unless people understand this they would vote against it as wasteful. A lot of projects, and the goal of the NSF, is to make it easy for people to understand these relationships and to respect the science, but I have a feeling that people won't go out of their way to even bother to try to understand it.

Anyways, /rant.
Math

'YouCut' Targets National Science Foundation Budget 760

Posted by timothy
from the conjugate-special-interest dept.
jamie writes "As some of you may have heard, the incoming Republican majority in Congress has a new initiative called YouCut, which lets ordinary Americans like me propose government programs for termination. So imagine how excited I was to learn that YouCut's first target — yes, its first target — was that notoriously bloated white elephant, the National Science Foundation."

Comment: The future (Score 1) 182

by kahizonaki (#33183600) Attached to: Google Testing an Airborne Camera Drone
Welcome to Google! "We know more about you than your own mother"
Search terms: | What is my girlfriend doing right now? |
...
...
Query returned: 2 results (3.793932 seconds for observation drones to move into position)
Result#1: Drone39103 (IR Camera, MMwavelength, NV Camera, Stereo camera) in position over 1722 Walnut St., Stalintown IN, residence of Ms. Wendy Smith, SSN 232-28-8821, google person number 399925800-1F
Result#2: Drone00192 (NV Camera, BFLaser) in position over 1722 Walnut St., Stalintown IN, residence of Ms. Wendy Smith, SSN 232-28-8821, google person number 399925800-1F
PS friendly message from your friendly google monitor: "ur gurlfrend is hawt lol! XOXO" END OF TRANSMISSION

SPONSORED ADVERTISEMENT: Vaporize your enemies! New, with google BigBrother! All drones are being outfitted with...

Comment: Re:Big fucking deal. (Score 1) 402

by kahizonaki (#32620886) Attached to: Tornado Scientists Butt Heads With Storm Chasers
The problem is that this argument can be applied to *any* situation, not just storm-chasing...

For instance, any time there is traffic/congestion, everybody involved is "causing problems for first responders to help people" (blocking the road by existing, changing lanes suddenly, etc.) and are "engaging in a dangerous activity" (driving in public). However, blaming the participants in a massively long traffic jam caused by e.g. an accident is not really justified, as you'd need to argue that they "have no right to be there" (out driving) because they're not there on "official" business. But, what's official business? Associated with their job, which contributes to local/global commerce? But, then there would be no place for leisure. So, assign everyone some leisure time. Then how do you decide which people are on the road on "official" leisure time versus clandestine leisure time? You would have to construct some national, nay, global schedule to determine when people are/aren't on leisure time, and require them to report it. And then everything could be controlled from a central control room with a large screen with a map of the world on it, and employ tens of thousands of analysts to optimize the oversight of all people's movement in the world. And then we wouldn't even need stop signs, everyone would be perfectly scheduled to move exactly at the right time so they'd properly interlace with everyone else and there would never be any inefficiency. Oops, sorry, I've just possibly had the most exciting dream as a computer scientists, and seem to have wet myself (in a most inappropriate manner).

Comment: Statistically relevant (Score 1) 446

by kahizonaki (#32553464) Attached to: Quant AI Picks Stocks Better Than Humans
In TFA it gives some numbers regarding the profitability of their system over the toy 5-week period. It is something like 8.5% whereas the next mutual fund is 6.5%. I don't see any language determining whether this is a statistically significant (or significant in any sense) difference at all -- I'm wondering if such at thing would even exist in these contexts. I would assume that any good prediction algorithm includes some stochasticity, so I wonder if this 2% difference can be relegated to noise or whether it is actually doing well. It seems like such a small amount after all (though, granted, 2% of 5 billion dollars is a nice chunk of change). But the other ones are doing something like 25% (in the world, I don't know the difference), so it certainly only seems to be doing mediocre at best...

Comment: Re:Perspective check (Score 1) 198

by kahizonaki (#31003402) Attached to: A Look Into the Chinese Hacker Underworld
The only problem is that the utility function to determine the values of X and Y strongly biases towards things more immediate in the future. Thus, doing something pleasurable now that you will pay for later is a decision that many make in a heartbeat, even if the value and consequence are objectively equal.

I remember reading somewhere (possibly some unpublished research?) that the utilities were weighted based on a hyperbolic function...i.e. non-linear. I can believe this, though I think it's more likely to have several "steps" in it as well. Just intuition.

Life is like an onion: you peel it off one layer at a time, and sometimes you weep. -- Carl Sandburg

Working...